Page 1


On the cover: Saturn and its moon Titan obtained using adaptive optics on the Gemini North telescope. For more details on this image and the related discovery of unexpected weather on Titan see the article by Henry Roe and Emily Schaller starting on page 7.



Director’s Message

27 “Wild Youth” of

Massive Galaxies

Doug Simons

Titan Monitoring

Henry Roe & Emily Schaller

10 Most Distant Known

Object in Universe

Edo Berger

13 Z~5 Star-forming

Galaxy with NIFS/IFU

Mark Swinbank

30 High-Z Band-Shuffling

with GMOS

Adam Muzzin & Howard Yee

34 AO Open-Loop Guiding

with Galactic Centers Richard McDermid, Davor Krajnović

& Michele Cappellari

39 Science Highlights

19 Mid-IR Observations

43 Exoplanet Imaging

Mark Westmoquette & Millicent Maier

of Seyfert Galaxies

Cristina Ramos Almeida & Chris Packham

23 Globulars in M31


Mariska Kriek & Pieter van Dokkum

17 IFU Wiki

Dougal Mackey, Annette Ferguson & Avon Huxor


Nancy A. Levenson

Christian Marois & Michael Liu

47 Time-Sharing


Dennis Crabtree & Tom Geballe

Exploring the Universe, Sharing its Wonders

52 Kyoto Science Meeting

Jean-René Roy, Atsuko Nitta & Nancy A. Levenson

55 On-site Observing

Dennis Crabtree

57 Planning for Future

Instruments Eric Tollestrup 60 MCAO Progress François Rigaut & Céline D’orgeville 62 Greening of Gemini Sarah Blanchard 67 Broadening Participation Neil Barker 70 GeminiFocus Reader’s Survey Peter Michaud 72 PIO Activities María Antonieta García, Peter Michaud

77 Introducing

Nancy A. Levenson Joy Pollard 78 Profile: François Rigaut María Antonieta García 82 Profile: Dolores Coulson Peter Michaud 86 Hawai‘i & Chile Nature Pictures by Staff

Joy Pollard & Manuel Paredes

Managing Editor, Peter Michaud Associate Editor, Carolyn Collins Petersen Proof Reader, Stephen James O’Meara Designer, Kirk Pu‘uohau-Pummill

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.



by Doug Simons Director, Gemini Observatory

Timing is Everything Figure 1.

This illustration (presented during the 1995 SPIE conference on astronomical instrumentation) is an early view of the planned Gemini cassegrain instrument cluster. A total of five beam feeds are provided around a rigid cube containing a highly articulated fold mirror and acquisition and guiding system. The design includes a calibration unit and an adaptive optics module that can feed any instrument. Electronics enclosures using air-liquid heat exchangers remove heat from the telescope environment. An infrared optimized up-looking port is preserved as well. This marvel of engineering serves as the “central nervous system” for each Gemini telescope and is fundamental to the success of our Target of Opportunity program.

Perhaps the only thing predictable about nature is its unpredictability. Of course, many natural events are highly predictable, such as the ebbing of tides and changing of seasons. But, as one tunnels deeper into natural phenomena, the vast majority of the universe appears unpredictable simply because it is so poorly understood. The splendor of the universe stems from not only its beauty but its capricious violence. Understanding natural chaotic phenomena is a field in which Gemini’s incredibly creative research community is growing adept, using Gemini as an observation platform. Back when Gemini was originally designed, however, we frankly never predicted these giant portals on the universe would be so effective at exploring astronomy’s time domain. This edition of GeminiFocus highlights several recent observations that beautifully illustrate how engineering decisions made long ago can sometimes, for one reason



or another, actually work together for a completely

underway before serious consideration was being given

different purpose today. The result is a powerful

to running a queue-based science operation model, and

new research tool for our community. This could

the term “Target of Opportunity” (ToO) was seldom,

be a lesson worth noting as the next generation of

if ever, heard in the bustling Gemini Project office

giant telescopes is designed, and similar tough design

while this system was being designed. Ironically, being

decisions are made to help them remain within budget

forced to consider “out of the box” solutions, due in

and seemingly limit their capabilities. The “moral to the

large part to the highly cost-constrained environment

story” is to not underestimate the discovery potential

in which Gemini was designed, bore the “fruit”

of an inventive community, powerful telescopes, and

hanging on the back of each Gemini telescope we

a target-rich sky to take astronomy in unexpected

enjoy today.


Another important design distinction at Gemini that

Like many observatories, the Gemini Observatory

is crucial to our ToO program’s success is, of course,

we know today bears, at best, a limited resemblance

the use of a queue-based operations model. Even in

to how it was first envisioned roughly two decades

this case, though, the rationale behind adopting a

ago. A tremendous amount of effort went into finding

queue-based model was only weakly linked to ToO’s

an optimal system performance within the Gemini

at its genesis. With telescopes that were engineered

8-meter Telescope Project’s finite budget. Some of the

to be site-limited in performance, and having several

effects of this design process are obvious, others less

instruments available concurrently, the dynamic range

so. A key early decision eliminated the Nasmyth foci

of the facility’s sensitivity and possibility of matching

at Gemini, leaving the telescopes with unconventional

observing programs with changing weather conditions

focal stations compared to our contemporary facilities

to maximize Gemini’s scientific product is what really

at Keck, Subaru, and the Very Large Telescope. Having

drove the queue concept behind the scenes. Finally,

instruments function exclusively in a Cassegrain

the last key ingredient in Gemini’s ToO system is

environment, compared to a more benign (and spacious)

simply the fact that Gemini—with twin telescopes—

Nasmyth focus, effectively transferred complexity and

has access to the entire sky; again, this originated in

risk from the telescopes to Gemini’s instruments—all

large part to provide ground-based back-up for space-

of which must fit within a 2,000-kilogram mass limit

based observations.

(900 kilograms for the adaptive optics (AO) systems) and limited space envelope, and function under any

The advantages in running a ToO program with

gravity vector.

8-meter telescopes on either side of the equator and

Recognizing these limitations, and out of concern

separated by ~10,000 kilometers was not a strong driver

at the time. In the end, it was a unique combination

for the notorious service and maintenance problems

of clever engineering, global coverage, and queue-

that frequent instrument changes cause, Gemini’s

based science operations that made Gemini the “ToO

engineers developed an innovative multi-instrument

machine” that is so powerful today. I wish those of

cluster capable of rapid remote reconfiguration while

us on the original project team could claim credit for

the instruments remain on the telescope for months

having the far-reaching scientific insight about research

at a time. It featured an AO system that could direct

trends in the 21st century and using that insight to

a corrected beam into any instrument mounted on

deliberately design an observatory that is so effective

the telescope. Merged with a clever calibration system

for ToO observations. However, the truth is—we just

(which replicates the telescope’s beam), an acquisition

got lucky.

and guidance (A&G) system (featuring three wavefront sensors), and on-instrument wavefront sensors (built

The ToO system now in place makes it possible to

into most instruments), the cassegrain cluster at

receive a signal at either Gemini telescope via the

Gemini packs an enormous range of capabilities and

Internet from a principal investigator (PI) anywhere

sophistication into a relatively small space. Figure 1

in the world. That signal, when handled by the night

shows the early (~ 1995) Cassegrain focal station concept at Gemini. Interestingly, this design was well

staff, allows them to seamlessly interrupt say, a NearInfrared Imager (NIRI) queue observation, slew the



telescope to the opposite side of the sky, configure

that captured the oldest photons ever recorded, sans

the Gemini Multi-Object Spectrograph (GMOS) for

those from the Cosmic Microwave Background, from

single-slit spectroscopy, and begin collecting photons from a ToO source. It often takes as little as 15 minutes

a gamma-ray burst that is a staggering z ~ 8.2 away (see Edo Berger’s article on page 10). Finally, it was

for a ToO trigger to yield science photons. Another

this system that made it possible to record fantastic

5-10 minutes after the observation is complete, the

mid-infrared images of the still-hot impact site on

anxious PI can retrieve the data from the Canadian

Jupiter after an unseen comet plowed into Jupiter’s

Astronomy Data Centre (CADC) and evaluate it

stormy atmosphere (see the Science Highlight on page

immediately, perhaps leading to another trigger using

39). We await with excitement the many targets of

a different instrument. It is this observing system

opportunity that will come within range of Gemini

that is responsible for the remarkable ToO science

in the years ahead. In the longer term, we foresee a

results found in this issue of GeminiFocus. It was this

rich ToO discovery horizon at Gemini because there

system that made it possible to observe the weather

are few large facilities better positioned to study the

on Titan using Altair and NIRI when photometry

fast-changing targets that will surely be revealed by the

from smaller telescopes indicated that interesting

next generation of synoptic survey telescopes.

meteorology was emerging on Saturn’s largest satellite


(see Henry Roe and Emily Schaller’s article on page

Doug Simons is Director of the Gemini Observatory. He can be

7, and cover image of this issue). It was this system

reached at:


ToO Science by Henry Roe & Emily Schaller

Titan’s Weather: Caught in the Act After several years of patiently waiting, in April 2008 Titan, Saturn’s largest moon, fell into a trap carefully set by our team of collaborators. On April 13, our spectral monitoring program at the NASA Infrared Telescope Facility (IRTF) indicated unusual cloud activity in Titan’s atmosphere. On the next evening, April 14, we triggered our Target of Opportunity (ToO) program at Gemini North to acquire high-resolution adaptive optics imaging to see what was happening. Our ToO sequence requires only 10-20 minutes of telescope time to image Titan using three filters that separate surface features, tropospheric methane clouds, and the global stratospheric haze. We trigger these ToO observations dozens of times per year based on cloud activity seen by one of our smaller telescope observing programs, cloud activity seen in previous Gemini data, the need for baseline data, and a variety of other reasons. The data are retrieved from the archive and reduced automatically every night, using a set of software scripts. For the first time in more than a year, these scripts failed when they encountered the April 14 data. A quick investigation that morning revealed that the new storm greatly outshone Titan; this had befuddled the reduction pipeline, leaving it unable to precisely locate Titan in the data. In our multiple years of observing Titan with Gemini, this was by far the largest storm we had ever seen. A few quick fixes to the code, and the pipeline



Figure 1.

using smaller telescopes (such as the 3-meter IRTF in

Near-infrared spectra of Titan from March 28, 2008, and April 13, 2008. Both spectra have approximately the same central longitude (227° W). But the spectrum from April 13 (red) shows increased flux in the transparent regions of Titan’s atmosphere (0.8, 0.95, 1.2, 1.6, and 2.0 microns), while staying the same brightness in the high-opacity regions. Because the surface reflection in these two spectra was essentially the same, the increased flux is due to tropospheric clouds. (Figure from Schaller et al. 2009.)

this case) to monitor cloud coverage on Titan each night, we can better and more efficiently allocate our precious and limited Gemini observing time to focus on Titan when it is most active. The April 2008 events also demonstrated why ground-based observations of Titan are so crucial, even with the Cassini spacecraft in orbit around Saturn. Cassini flies by Titan every few weeks, but has limited ability to observe that world in between those flybys. In the spring of 2008, Cassini flew by Titan on March 25 and then did not happily reduced these new data. (See the image

visit Titan again until May 12. In Cassini images from

labeled “2008-04-14 UT” in Figure 2 on page 9.)

the later flyby, a few, faint strange-looking clouds remain. However, for the most part, Cassini entirely

Why were these observations so important? Titan

missed the events of the previous month.

resembles Earth because many of the processes active on its surface, and in its atmosphere, are similar

Most importantly, the events of April 2008 gave

on both worlds. While the materials involved are

new insight into how and where clouds form on

somewhat alien to us on Earth but perfectly normal

Titan. That initial storm of April 13-16, 2009, was

for Titan, the processes are completely familiar. For

convective and violent enough to kick off waves in the

instance, Titan’s clouds, rain, and surface liquids

atmosphere, just as a dropped stone generates waves

are all made of methane rather than water, but the

on the surface of a pond. These ripples, technically

processes of methane cloud formation, rainfall, and

known as Rossby waves, then traveled throughout

evaporation from lakes all have Earthly analogs. Most

Titan’s southern hemisphere and triggered cloud

water-based hydrologic or meteorologic processes

formation elsewhere, including at the south pole and

here on Earth have directly analogous methane-based

in the equatorial regions. On April 28, the wave had

processes on Titan. Like Earth, both Mars and Venus

circled Titan and caught up to the original storm,

are terrestrial planets and are closer in size to Earth

triggering new cloud formation and a brightening of

than Titan, but from the point of view of planetary

the storm.

processes on the surface and in the atmosphere, Titan stands out as the most similar body to Earth in our

These were the first observations of wave activity

Solar System.

in Titan’s atmosphere and demonstrated convincingly that a strong-enough punch to the atmosphere in

Titan has seasons for the same reason Earth does:

one location can trigger cloud formation anywhere

its axis of rotation is tilted 26.4 degrees, just three

in the same hemisphere on Titan. This overturned a

more degrees than Earth’s tilt. However, Titan’s year

recently developed understanding of Titan suggesting

is much longer. One Titan year lasts 30 Earth years.

that its equatorial region is a dry desert and should never see clouds or rainfall, while the polar regions

We’ve only had the technology to observe Titan’s

should be wet in early summer and dry the rest of

methane weather directly for about a decade, and

the year. These observations demonstrate that clouds,

our campaign to study the weather in detail began

and presumably rainfall, do occur at the equator

only about five years ago. Titan’s northern vernal

and can occur out of season at the poles. Further,

equinox occurred in August 2009. Imagine trying to

this helps solve the mystery of what process could

understand Earth’s complicated weather from only

have formed the dry streambeds and channels seen

having seen January, February, and part of March,

at Titan’s equator by Cassini and its lander probe

and you can see why we are planning to continue to

Huygens. Researchers had begun to invoke the

observe Titan for decades to come.

existence of underground springs, geysers, or other even stranger geologic features, but these observations


The events of April 2008 demonstrated the value of

show that methane rain showers are the likely source

a multi-telescope observing approach for Titan. By

of the liquid that cut those streambeds.


Figure 2.

These images from Gemini North were taken using a filter (H21-0, 2.12 microns) that obscures Titan’s surface, but highlights the variable bright clouds and more static high stratospheric haze. On March 27, 2008, there were essentially no clouds present. The central longitude is indicated above each image. The green box indicates the initial location of the large storm. The original storm rotates to the night- side of Titan after a few days with Titan’s 16-Earth-day rotational period.

The biggest remaining mystery from April 2008 is:

For further information see:

What was the original trigger? What could have

Schaller, E. L., Roe, H. G., Schneider, T., & Brown,

punched the atmosphere so hard as to set off one of

M. E., 2009, Nature, 460, 873.

the biggest storms yet seen on Titan? The most likely two possibilities are seasonal atmospheric dynamics

Henry Roe is an astronomer at the Lowell Observatory in

or some type of geologic activity on the surface. The

Flagstaff, Arizona. He can be reached at:

atmospheric dynamics possibility requires just the right combination of winds shifting around on Titan,

Emily Schaller is an astronomer at the University of Arizona

which could happen. From Cassini images there is

(however, much of this work was done while she was at the

strong evidence that Titan’s surface is still active, with

Institute for Astronomy at the University of Hawai‘i). She can

cryovolcanoes, mountain formation, and erosion, and

be reached at:

possibly even small-scale plate tectonics. Any of these processes could release enough subsurface methane into the atmosphere to kick off a storm of the size we saw. Curiously, the surface region where the storm initiated (shown by a green box in Figure 2) remained bright for weeks. With further observations of Titan over the next few years, we’ll be able to distinguish between these scenarios as our network of telescopes and observing programs capture more of Titan’s methane weather.



ToO Science by Edo Berger

The Most Distant Known Object in the Universe Figure 1.

The fading infrared afterglow of gamma-ray burst GRB 090423 appears in the center of this falsecolor image taken with the Gemini North telescope in Hawai‘i. This burst is the farthest cosmic object/event yet seen.

One of the most fundamental questions facing 2 1 st- c e n t u r y a s t ro n o my  i s how the universe evolved from the simple and smooth remnant of the Big Bang to




complex interplay of stars, galaxies, and clusters that we see around us today. A key piece of this puzzle is the formation of the first stars and galaxies, and their inevitable role in reshaping the universe around them. This is called the process of cosmic reionization. On April 23, 2009, I participated in a discovery that is sure to shed light on this cosmological inquiry. We found evidence for the most distant known object in the universe—a type of explosion known as a gammaray burst (GRB). Our observations showed that this burst, dubbed GRB 090423, took place when the universe



Figure 2.

The redshift distribution of gamma-ray bursts, including the record-breaking GRB 090423 at z = 8.3. Also shown are the highest redshift galaxy (z = 6.96) and quasar (z = 6.43).

was only 630 million years old—a mere 4.6% of its

been driving a race between several astronomical

current age. The discovery of this extraordinary

communities: those studying bright quasars, faint

event pushes our view of the universe earlier by

galaxies, and most recently, GRBs. Starting with

150 million years of cosmic evolution, bringing us

their discovery in the 1960s, and up until the mid-

ever closer to the very first generation of stars and

1990s, quasars have dominated the race. These


objects, powered by billion-solar-mass black holes, are extremely luminous, but also exceedingly rare.

We could detect GRB 090423 and determine its

The advent of the Hubble Space Telescope and

redshift at such an extreme distance for two

large ground-based telescopes in the 1990s propelled

important reasons. First, GRBs and their long-

distant galaxies into the forefront, and they have

wavelength afterglows are the most luminous objects

since been competing neck-and-neck with the distant

in the universe. In this particular case, the afterglow

quasars. Indeed, before our April 23 discovery, the

peaked at a K-band AB magnitude of about 20, at least

most-distant known object in the universe was a

10 magnitudes brighter than the expected luminosity of typical galaxies at a comparable redshift! Second,







Opportunity capability to determine a photometric redshift of z = 7.6-9.2 within two hours of the burst discovery (see Figure 2). This value was later refined spectroscopically to z = 8.3. How does this record-breaking burst fit within the broader distribution of high-redshift objects? Over the past several decades, the quest to study the highredshift universe and the process of reionization has


Figure 3.

Near-infrared observations of GRB 090423 in the Y-, J-, H-, and K-bands. The dashed red line indicates spectral flux density in the absence of neutral hydrogen absorption at high redshift, while the solid red line includes the effect of absorption at the redshifted wavelength of the Lyman-alpha line. The blue line indicates the effect of dust extinction, which clearly cannot explain the sharp cut-off at about 1.2 microns.


small galaxy dubbed IOK-1, at a redshift of 6.96.

be explained by absorption in the high redshift IGM

Only slightly behind that was a quasar called

and/or by neutral hydrogen within the host galaxy.

CFHQS J2329-0301, which lies at a redshift of 6.43.

In either case, the redshift was unambiguous.

GRBs entered the race much later, but have quickly taken the lead.

Our discovery of GRB 090423 at a record redshift of about 8.3 opens new distant horizons. It proves that

Starting with the first distance measurements for

stars were already dying at a time when the universe

GRBs in 1997, they have quickly climbed the distance

was only 630 million years old. Equally important,

ladder, reaching out to a respectable redshift of 6.29

we now know that when the first stars died, at

in 2005 and overtaking the quasars with a redshift

least some of them exploded as bright GRBs. This

of 6.7 in 2008 (Figure 2).

gives us a promising way to pinpoint their location and to find increasingly more distant objects in

The quest for high-redshift galaxies and quasars has

the future. Although new observations with the

been stalled at z ~ 7 for physical reasons. Beyond this redshift, the optical emission is fully absorbed by

WFC3 instrument on Hubble have now uncovered

neutral hydrogen in the intergalactic medium (IGM),

potential z ~ 8 galaxies, it is unlikely that these will be spectroscopically confirmed until well into the

and searches require extensive infrared observations.

next decade.

Wide-field infrared surveys are required to find the bright but exceedingly rare z > 7 quasars, while

The Swift satellite, on the other hand, can certainly

deep narrow-field observations are more conducive

detect extremely distant bursts, and we are poised

to finding the more numerous, but much fainter,

to determine their distances using telescopes such

z > 7 galaxies. Spectroscopic confirmation of faint

as Gemini.

candidate z > 7 galaxies is currently another bottleneck. It is unlikely to be resolved before the launch of

Ultimately, our desire to find the most distant

NASA’s James Webb Space Telescope around 2015.

GRBs goes beyond just bragging rights and redshift

In all cases, the signature of a high-redshift origin

records. Because they are so bright in infrared light,

is a sharp cut-off at the redshifted wavelength of

the bursts can be used to momentarily illuminate

the Lyman-alpha line induced by the IGM. This

their otherwise extremely faint galaxies and the

requires deep limits in the optical band and high

increasingly neutral IGM. Infrared spectroscopy will

signal-to-noise ratio detections in the infrared bands

thus allow us to measure the metallicities of the

to avoid confusion with more mundane processes

highest redshift galaxies, in much the same way that

effects, such as dust extinction and intrinsic spectral

we have done at z<6 over the past several years.

features in galaxy or quasar spectra.

Equally important, such spectroscopic observations will allow us to precisely measure the neutral fraction

In the case of GRBs, the signature of a high redshift

of the IGM well into the process of reionization.

is similar, but its detectability is aided by the fact that the afterglow spectrum is due to featureless

For more information see:

synchrotron emission. Thus, the only possible source

Tanvir, N., Fox, D. B., Levan, A. J., Berger, E., et al.

of confusion is dust extinction. In the case of GRB

2009, Nature in press; arXiv:0906.1577

090423, deep optical limits less than 20 minutes after

Bouwens, R. J., et al. 2009, Submitted to ApJL;

the burst discovery and a detection in the K-band


suggested either a high-redshift origin or significant

Iye, M., et al. 2006, Nature, 443, 186

dust extinction. Armed with this knowledge we

Willott, C. J., et al. 2003, ApJ, 587, L15

immediately slewed the Gemini North 8-meter telescope to the burst’s location and obtained Y-,

Edo Berger is an assistant professor of astronomy at Harvard

J- and H-band observations with the Near-Infrared

University. He can be reached at:

Imager (NIRI). The results of the observations are shown in Figure 3 and a color image is shown in Figure 1. The NIRI observations clearly demonstrated a sharp cut-off at about 1.2 microns that could only



IFU Science by Mark Swinbank

Studying a Lensed Z ~ 5 Star-forming Galaxy with NIFS/IFU A key extragalactic science driver for the justification of the next generation of telescopes, such as the Thirty Meter and Extremely Large Telescopes, is to detail the internal properties of high-redshift galaxies on scales comparable to HII regions. This will allow us to understand how and why galaxies in the distant universe are so much more efficient at forming stars than they are today. The study of the internal structure of typical high-redshift galaxies requires both superb sensitivity and high resolution. While present-day laser-guide-star adaptive-optics-fed (LGS AO) integral field units (IFUs) deliver resolved data for z ~ 1.5 - 3 galaxies on about kiloparsec scales, such observations only provide a few independent resolution elements for representative high-redshift sources typically less than 2 kiloparsecs across. Fortunately, we need not wait for the next generation of larger telescopes to obtain a much-improved spatial sampling. One of the most striking natural phenomena seen in the universe is strong gravitational lensing by galaxy clusters, and these massive clusters act as natural telescopes. They amplify the images of distant galaxies which serendipitously lie behind them, boosting their fluxes and apparent sizes by factors of 10 to 30. By combining gravitational lensing with IFU spectroscopy we can therefore probe the dynamics and distribution of star formation and the interplay between star formation and gas dynamics in much greater detail than otherwise possible. Indeed, the spatial resolution can even reach 100 parsecs, which begins to resolve individual HII regions. This feat would otherwise be impossible without the light grasp and resolution of a 30- to 40-meter-class telescope. One of the most spectacular gravitational arcs is the famous lensed galaxy behind MS 1358+62 at z = 4.92 (Figure 1, see next page). Discovered during a Hubble Space Telescope (HST) imaging campaign in 1997, this galaxy offers a nearly unique opportunity to study the internal processes that drive galaxy formation



Figure 1.

A true-color HST ACS VRI-band image of the lensed z = 4.92 galaxy behind MS 1358+62 with the z = 4.92 critical curve from our best-fit lens model overlaid. The HST images clearly show the multiply imaged galaxy as two mirror images which are folded about the critical curve (red line resolved into two images of the background galaxy). The morphology of the galaxy is clearly dominated by up to six starforming regions surrounded by a diffuse halo. The upper middle panel shows the white light (wavelength collapsed) image of the [OII] emission from the IFU; the boxes labeled 0-4, denote the regions from which spectra were extracted (see Figure 3).

in the early universe—likely at an epoch when a

data allow us to spatially resolve the distribution

galaxy forms its first generation of stars. With multi-

of star formation and dynamics. As Figure 1 also

wavelength imaging from HST’s Advanced Camera

reveals, the nebular emission-line map closely

for Surveys (ACS) and Near-Infrared Camera and

follows the rest-frame ultraviolet (UV) morphology,

Multi-Object Spectrometer (NICMOS), as well as

and by collapsing the data cube from the brightest

the Spitzer Space Telescope, this well-studied galaxy

regions, we derive a peak-to-peak velocity gradient

cluster has a wealth of arcs and arclets that allow

of 180 kilometers per second across 4 kiloparsecs,

for a detailed map of the mass distribution within the galaxy cluster to be constructed. In Figure 2 we

suggesting a dynamical mass of Mdyn ~ 3 × 109 MSun For comparison, the stellar mass (estimated from the

show the source-plane reconstruction of the galaxy

rest-frame UV-to-K-band spectral energy distribution)

image and use this to derive a luminosity-weighted

is ~ 109 MSun.

amplification factor of 12.5 and a source-plane radius of ~ 2 kiloparsecs. As Figure 2 also shows, the source-

This dynamical mass is an order of magnitude

plane morphology is dominated by approximately

smaller than the median Lyman-break galaxy mass

five star-forming regions.

at z ~ 3 for which masses have been measured using similar techniques. However, it is comparable to

In order to investigate the internal properties of the

the only other star-forming galaxy at this redshift

galaxy in detail, we used the Near-Infrared Integral

(RCS 0224-002 arc at z = 4.88) with a dynamical mass

Field Spectrometer (NIFS) IFU to map the nebular

measured in a similar way.

[OII] 3727 emission line doublet from the galaxy. Our



Figure 2.

outflow is young (less than 10 million years) and has yet to decouple and escape from the galaxy disk. Using a combination of the instantaneous starformation rate, the P-Cygni Lyman-alpha emission line profile, and the UV-ISM absorption line strengths, we estimate the kinetic energy in the wind 54

(~ 10 erg) and mass loading in the wind (greater

than 3 × 105 MSun). In comparison, the kinetic energy provided by the collective effects of stellar winds

and supernovae explosions over 15 million years (KE 57 ~ 5 × 10 erg) is easily enough to drive the wind. The low inferred mass loading and kinetic energy

in the wind is in stark contrast to the z = 4.88 galaxy behind the lensing cluster RCS 0224-002 (the only The spatially resolved velocity gradient seen in the galaxy disk can be compared directly to that measured from the Lyman-alpha and ultraviolet

A true-color HST ACS VRI-band reconstruction of the lensed galaxy. The amplification of the galaxy is a factor 12.5+/-2.0. In the source plane, the galaxy has a spatial extent of ~4 kiloparsecs and comprises at least five discrete starforming clumps. The largest of these is only marginally resolved, with a restframe optical half light radii of ~ 200 parsecs.

other galaxy at these early times for which similar measurements have been made). In this system, a large-scale (more than 30 kiloparsec) bipolar outflow

(UV)-interstellar medium (ISM) velocity structure.

Figure 3.

(Top) A onedimensional spectrum of the five star-forming regions within the z = 4.92 galaxy from the NIFS IFU observations. In all panels the position of the [OII] 3726.8, 3728.9 doublet is at a fixed redshift of z = 4.9296. The final panel shows the one-dimensional spectrum of the z = 4.92 galaxy around the Lymanalpha emission. (Bottom) The extracted, onedimensional velocity gradient along the long axis of the galaxy (source plane).

Franx et al. (1997) discuss the rest-frame UV spectral properties of this galaxy in detail. In particular, the UV-ISM absorption lines show velocity variations on the order of 200 kilometers per second along the arc with the [SiII] 1260 absorption line systematically blueshifted






emission, and an asymmetric Lyman-alpha emission line with a red tail. These spectral features are naturally explained by an outflow model in which the blue side of the Lymanalpha line has been absorbed by outflowing neutral HI. The description of the line profile is typical of the emission profiles seen in other Lyman-break galaxies. Indeed, velocity offsets between the nebular emission lines (such as Hα) and UV-ISM emission/ absorption lines are now common in high-redshift star-forming galaxies, and are usually interpreted as evidence for a large-scale starburst-driven outflow. As Figure 3 shows, within the z = 4.92 galaxy, the nebular emission line gradient from the [OII] emission (~ 180 kilometers per second across ~ 4 kiloparsecs) is mirrored (but systematically offset) from the Lyman-alpha and rest-frame UV-ISM lines, suggesting that this galaxy is surrounded by a galactic-scale outflow. Since the velocity gradient observed in the nebular emission is mirrored (but systematically offset) from the Lyman-alpha emission and UV-ISM absorption, this also suggests that the

has a high mass loading (greater than three times the star-formation rate) and is escaping at speeds of up to 500 kilometers per second. The strong contrast between the only two galaxies where such studies have been made at these early times suggests strong diversity in the outflow energies of young galaxies at high redshifts. This clearly illustrates the need for more targets and follow-ups to test the impact and ubiquity of outflows in the early universe.



Figure 4.

to draw strong conclusions since there is little

Correlation between size and starformation rates from H-alpha for HII regions in local galaxies (including the largest local HII region 30 Doradus in the LMC and the nearby galaxies IIZw40 and the Antennae). The two HII regions in MS1358+62 arc are shown by the solid red points. The dashed line shows a fit to the HII regions in nearby spiral galaxies whilst the dotted line is the same but shifted in luminosity at a fixed size by a factor of 100.

observational evidence for strong variations in the IMF (especially at high redshift). However, partially resolved spectroscopy of the nebular emission lines on scales comparable to HII regions would provide crucial diagnostics of the physics of star formation in the young universe. Indeed, the mechanical energy input from a higher fraction of OB stars might explain the higher turbulent speeds recently observed in primitive disks at z ~ 3 (eg. Stark et al. 2008). Next, we estimate the sizes, masses, and starformation rates of the two brightest star-forming HII regions in the galaxy, which are all well resolved in our data. Using the optical, near-infrared, and IFU data, we estimate that the two brightest HII regions are on the order 100 - 200 parsecs across and have line widths (from the [OII] emission) of ~ 80 - 90 kilometers per second. This suggests that the largest HII regions have masses on the order of 2 - 3 × 108 MSun A comparison to HII regions from galaxies in the local universe shows that these are comparable in mass and density to most local, massive star-forming HII regions. However, the starformation rate for these HII regions (individually) is > 100 times higher than typically found at a fixed size (Figure 4). Such high star-forming rates (for a fixed size) have been observed locally in the most extreme systems, such as (eg. 30 Doradus in the LMC). Thus, the HII regions within this z = 4.92 galaxy appear to be scaled-up versions of the most extreme regions observed in the local universe, yet observed when the universe was only about one billion years old. Could the increased luminosity within HII regions reflect real differences in the mode of star formation within massive star-forming complexes? Recent 7

models suggest that young, massive (> 10  MSun), starforming clusters are optically thick to far-infrared radiation, resulting in high gas temperatures, and, hence, higher Jeans masses. Indeed, for a 108 MSun star cluster (as observed in the z  = 4.92 arc) the predicted Jeans mass is ~ 12 MSun. This “top-heavy” initial mass function (IMF) has the effect of increasing the fraction of OB stars per HII region and hence, increases the light-to-mass ratio. Although such a model could provide a fit to the data, it is dangerous



Overall, these observations provide unique insights into the galaxy dynamics and the intensity and distribution of star formation, as well as the interaction between star formation and outflow energetics within a young galaxy seen less than 1 billion years after the Big Bang, on scales of just 200 parsecs. The large number of high-redshift gravitationally lensed galaxies identified by HST, combined with adaptive optics assisted integral field spectrographs on 8- to 10-meter telescopes, should finally begin to make such critical observational studies commonplace and allow us to test the route by which early systems assemble their stellar masses, their modes of star formation, and how they ultimately develop into galaxies like the Milky Way. The international university-based research team led by the author included: Tracy Webb (McGill), Johan Richard (Durham), Richard Bower (Durham), Richard Ellis (Caltech), Garth Illingworth (UCO), Tucker Jones (Caltech) Mariska Kriek (UCO), Ian Smail (Durham), Dan Stark (Cambridge), Pieter Van Dokkum (Yale) For more information see: “A Spatially Resolved Map of the Kinematics, StarFormation and Stellar Mass Assembly in a StarForming Galaxy at z = 4.9” in press, MNRAS. Mark Swinbank is an astronomer at the University of Durham (UK). He can be reached at:

IFU Resources by Mark Westmoquette & Millicent Maier

The Integral Field Spectroscopy User’s Wiki The field of integral field spectroscopy (IFS) is now well developed, with IFS instruments installed on all the main optical telescope facilities around the world. Gemini Observatory has three of the most powerful high spatial and spectral resolution integral field units (IFUs): The Gemini Multi-Object Spectrograph (GMOS) installed at both locations and the Near-infrared Integral Field Spectrometer (NIFS) at Gemini North. However, although excellent work based on IFS data is being published by many groups, IFS continues to be avoided by large sections of the astronomical community due to perceived difficulties with data handling, reduction, and analysis. There is no doubt that dealing with IFS data is more complicated than simple imaging or long-slit spectroscopy, but many of the problems that arise could easily be avoided by benefiting from the experience and knowledge of others. In our experience, a lack of information sharing has forced many groups to come up with their own independent solutions to data cube manipulation, visualization, and analysis (in a sense “re-inventing the wheel” each time), and because of the effort needed to do this, the tools developed are not automatically made publicly available. Therefore, we thought it valuable to have a central repository of information: tips, codes, tools, references, etc., regarding the whole subject of IFS, which is accessible and editable by the whole community. To this end, we have set up the IFS wiki, which we hope will become this repository:



Figure 1.

experience with particular instrumental quirks, then

A screenshot of the IFS wiki’s welcome page.

please add this to the site. This type of information is of essential use to all the community. The site will be regularly edited/moderated for structure and English to keep it as useable and clear as possible. So contributors who are worried about their English need not worry. The originators and maintainers of this site are After spending some time building up a solid foundation of content, the wiki was opened (made fully editable) to the community on May 18th, 2009. As of now, the wiki covers the following topics: current and future integral field spectrographs (including all those currently on the Gemini telescopes); observing techniques and observation planning; data reduction, including basic overview of the procedures for the different types of IFS instruments, and more advanced/specific tasks like mosaicking or differential atmospheric refraction (DAR) correction; and analysis techniques, from visualizing, to line fitting, to source extraction.

Figure 2.

The main techniques for achieving integral field spectroscopy. Adapted from AllingtonSmith, Content and Haynes (1998). The Lenslets and Fibres design in the central row is most similar to the GMOSIFU set-up. The Slicer design on the bottom row is most similar to NIFS’s.

If you count yourself in the “less experienced” category, and you want to know more about IFS, then please visit the site—make it your first port of call if you have any questions or issues with IFS data. However, if you’ve had any experience with IFS, please think about contributing your knowledge to the wiki. Almost all of the Gemini staff who work with these instruments have already contributed their expertise to specific sections. Finally, for the site to develop its full potential, we need the community to make continued contributions: if something is missing or wrong, if you have a useful hint or piece of code that has really helped you with your data, or you have



currently Dr. Katrina Exter (KULeuven, Belgium) and Dr. Mark Westmoquette (UCL, UK). Mark Westmoquette is a postdoctoral fellow at the Physics and Astronomy Department, University College London, UK. He can be reached at: Millicent Maier is a science fellow at Gemini Observatory located at Gemini South. She can be reached at:

Mid-IR Science by Cristina Ramos Almeida & Chris Packham

Modelling Nuclear Infrared Emission of Seyfert Galaxies on Parsec Scales The unified model of active galactic nuclei (AGN) is predicated on the existence of a dusty toroidal structure surrounding the object’s central engine. Accretion onto a supermassive black hole is commonly thought to produce the enormous luminosity observed in these objects. As the spatial resolution afforded by existing telescopes is typically insufficient to resolve the torus, several studies have attempted to model the infrared (IR) emission of the torus, permitting investigations of different geometries, sizes, structures, and orientations to our line of sight. However, previous modelling was tested against relatively poor spatial resolution data (angular resolution larger than ~ 1 arcsecond), which include a substantial amount of host galaxy “contamination”. Hence, the optimal way to infer the torus’s properties is to model the IR emission with the high angular resolution afforded by the 8-meter class of telescopes. For this purpose, we have constructed subarcsecond resolution IR spectral energy distributions (SEDs) for 18 nearby Seyfert galaxies. The data were predominantly obtained from both Gemini North and South telescopes in the mid-infrared (MIR). In addition, near-infrared (NIR) data of similar angular resolution have been compiled from the literature. The precise torus geometry, structure, and size have been a matter of discussion for some time. Pioneering work assumed a homogeneous distribution of dust, as there was no other observational evidence and it simplified the modelling. In recent years, in order to account for new observational constraints and physical improvements over homogeneous torus models, a clumpy distribution of dust in a toroidal shape is becoming widely accepted. Figure 1 (next page) shows a graphical representation of the clumpy torus. Inhomogeneous models are making



significant progress in accounting for the MIR emission

in December 1998, at the 4-meter Blanco Telescope at

of AGNs. A fundamental difference between clumpy

Cerro Tololo Inter-American Observatory (CTIO),

and smooth density distributions of dust is that the

and in May 2001, at the Gemini North telescope.

former implies that both directly illuminated and

Another set of observations was performed with

shadowed cloud faces may exist at different distances

the MIR camera/spectrograph T-ReCS at the Gemini

from the central engine. Thus, the dust temperature

South telescope, and the last one with MICHELLE

is not a function of the radius only, as was the case

at Gemini North. A summary of the observations,

in the homogeneous models, since illuminated and

together with the afforded resolutions for each galaxy,

shadowed clouds contribute to the IR emission from

are also presented in Table 1.

all viewing angles. This leads to a small sized torus, where the long-wave IR emission can even originate

To construct well-sampled IR SEDs, we compiled

close to the central engine.

NIR nuclear fluxes at similarly high spatial resolution from the literature. These fluxes correspond to the

Figure 1.

A schematic of the clumpy torus. The radial extent of the torus is defined by the outer radius (R0) and the dust sublimation radius (Rd). All of the clouds are supposed to have the same optical depth (tV), and σ characterizes the width of the angular distribution. The number of cloud encounters is a function of the viewing angle, i (Nenkova et al., 2008).

observed emission from the nuclear region of the galaxies (unresolved component), with typical spatial scales of the same order as those of our nuclear MIR fluxes.

The IR SED and Observational Constraints Crucial observational constraints for torus modelling arise from the shape of the IR SED of Seyfert galaxies. The bulk of the torus emission is concentrated in this wavelength range, as the torus blackbody emission


peaks at the MIR wavelengths (7 – 26 μm). When large aperture data, such as those from Infrared

G r o u n d - b a s e d  M I R  h i g h - a n g u l a r  r e s o l u t i o n

Space Observatory (ISO), Spitzer, or the Infrared

observations of 18 nearby active galaxies were carried

Astronomical Satellite (IRAS), are employed for

out over the past several years for a variety of science

constructing SEDs, the IR fluxes are a mixture of

cases. By pooling these data sets, we make use of this

AGN plus host galaxy emission. From the comparison

archive of data here for the entirely different science

between our IR SEDs and large aperture data from

case discussed above. Most of these sources are Type

2MASS, ISO, and IRAS, we confirm that the high spatial

2 Seyferts, but the sample also includes two Seyfert 1.9

resolution MIR measurements provide a spectral shape

(Sy1.9), one Seyfert 1.8 (Sy1.8), two Seyfert 1.5 (Sy1.5),

of the SEDs that is substantially different from that of

and one Sy1 galaxy (see Table 1).

large aperture data SEDs (for an example see Figure 2). Thus, our nuclear measurements allow us to better

Table 1.

characterize the torus emission and consequently use

Summary of midinfrared observations for the total sample of Seyfert galaxies.

torus models to constrain the distribution of dust in the immediate AGN vicinity. To derive general properties of Sy2 galaxies, which constitute the majority of our sample, we constructed an average Sy2 SED, considering only the highest angular resolution data to avoid as much as possible the stellar contamination. This average Sy2 template was used to compare with the individual SEDs of


The first set of observations was obtained with the

the Sy2 analyzed in this work, and also with the

University of Florida MIR camera/spectrometer OSCIR,

intermediate-type Seyferts (see Figure 3).


Figure 2.

For the case of the Sy2 galaxies, we find a relatively low number of clouds (N0 = [5, 15]) and large values of the torus width (σ = [50°, 75°]) from the fits. High values of the inclination angle of the torus are more probable (i = [40°, 90°]) and the optical depth per cloud is generally lower than tV = 100. For the intermediate-type Seyferts we find smaller number of clouds than for the Sy2 (N0 = [1, 7]). Low values of

σ are preferred from the fits of Sy1.8 and Sy1.9 (σ

= [25°, 50°]) and even lower for Sy1.5 (σ < 35°). We require direct (though extinguished) emission of the AGN to reproduce the near-IR excess observed in the The slope of the IR SED is, in general, correlated with the Seyfert type. Sy2 show steeper SEDs, and intermediate-type Seyferts are flatter. Seyferts 1.8 and 1.9 present intermediate values of the IR slope and the NIR to MIR flux ratio between Sy2 and Sy1.5. This ratio increases as the hot dust emission becomes important, giving an indication of the torus geometry and inclination. However, we find a range of spectral shapes among the Sy2 galaxies, and some intermediatetype SEDs have the same slopes as the Sy2. This cannot be reconciled with the predictions of smooth torus models, since large optical depth homogeneous tori strictly predict steep SEDs for Sy2 and flat SEDs for Sy1. We do not observe such a strong dichotomy, and

SEDs of Sy1.5. Views of Sy2 are more inclined than those of the Sy1.5 galaxies. More importantly, the large values of N0 and σ resulting from the Sy2 modelling suggest that their central engines are blocked from a direct view along more lines of sight than are those

A comparison between the IR high angular resolution data employed in this work (filled circles) and low resolution data from 2MASS, ISO, and IRAS (taken from the NASA/ IPAC Extragalactic Database – NED; star symbols) for the Circinus galaxy. Note the difference in steepness between them and the reduced MIR T-ReCS fluxes, compared with those from IRAS at approximately the same wavelengths.

of the intermediate-type Seyferts, which would have clearer views of their AGN engines. This would imply that the observed differences between Type 1 and Type 2 AGNs would not be due to orientation effects only, but to different covering factors in their tori. However, due to the limited size of the analyzed sample, these differences are not statistically significant, and a larger sample is needed to confirm whether Sy1 and Sy2 tori are intrinsically different.

so we pursue the clumpy torus models to reproduce the observed SEDs.

Figure 3.

Top: Observed IR SEDs for the seven Sy2 galaxies (in color and with different symbols) used for the construction of the average template (solid black). The SEDs have been normalized at 8.8 μm, and the average SED has been shifted in the Y-axis for clarity. Bottom: Comparison between the average Sy2 SED (solid black) and the intermediate-type Seyfert SEDs. Note the difference in steepness.

Model Results We fit the observed SEDs with the clumpy torus models described by Nenkova and collaborators, using the BAYESCLUMPY tool. From Figure 4 (next page) it is clear that the Circinus galaxy data (eight photometric data points) provide sufficient information to the code to constrain the model parameters. The clumpy models successfully reproduce the infrared SEDs of the Seyfert galaxies analyzed here. The models accommodate the range of IR slopes observed in the Sy2, while providing optically thick obscuration along the line of sight. The IR SED fitting does not constrain the radial extent of the torus, and this parameter remains uncorrelated with the other parameters that define the models. Because the outer torus contains the coolest material, high angular resolution measurements at wavelengths longer than 15 μm are needed to reveal significant variations in the torus size.



Figure 4.

Ramos, (Instituto de Astrofísica de Canarias, Tenerife,

High spatial resolution IR SED of the Circinus galaxy fitted with the clumpy torus models. Solid and dashed lines are the best fitting model and that computed with the median of each of the six parameters that describe the models. The shaded region indicates the range of models compatible with a 68% confidence interval around the median. T-ReCS N and Q images (8.8 and 18.3 microns, respectively) are also shown.

Spain); James T. Radomski, (Gemini Observatory); Scott Fisher, (Gemini Observatory); Charles M. Telesco, (University of Florida). For more information see: Alonso-Herrero, A., et al. 2006, ApJ, 652, L83 Asensio Ramos, A. & Ramos Almeida, C. 2009, ApJ, 696, 2095 Antonucci, R. R. J. 1993, ARA&A, 31, 473 Díaz-Santos, T., et al. 2009, ApJ, submitted Nenkova, M., Sirocky, M. M., Nikutta, R., Ivezic, Z., Only through parsec-scale observations of several

and Elitzur, M. 2008, ApJ, 685, 160

AGNs, such as those presented here, can the torus

Packham, C., et al. 2005, ApJ, 618, L17

geometry and properties be well constrained. Our

Ramos Almeida, C., et al. 2009, ApJ, 702, 1127

results strongly support the clumpy torus models and

Urry, C. M. & Padovani, P. 1995, PASP, 107, 803

provide constraints on the structure of the torus. To further these studies, we plan more similar observations

Cristina Ramos Almeida, recently at Instituto de Astrofísica

(including 10-μm spectroscopy) to provide a more

de Canarias (Tenerife, Spain) where this work was completed,

detailed characterization of torus properties.

is now a postdoctoral research associate in AGN & galaxy evolution at the Physics and Astronomy Department of the

Our collaborators in this study are: Nancy Levenson,

University of Sheffield (UK). She can be reached at:

(Gemini Observatory); Jose Miguel Rodríguez Espinosa,

(Instituto de Astrofísica de Canarias; Tenerife, Spain);


Almudena Alonso Herrero, (Instituto de Estructura

Chris Packham is an associate scientist at the University of

de la Materia, CSIC, Madrid, Spain); Andrés Asensio

Florida. He can be reached at:


High-Res Imaging by Dougal Mackey, Annette Ferguson & Avon Huxor

Unveiling Remote Globular Clusters in Andromeda G l o b u l a r  c l u s t e r s  a r e  w i d e l y recognized as extremely useful tracers of galaxy formation and evolution. Typically, they are compact and luminous objects that can be observed out to large distances. Often they are among the oldest components of any given galaxy. Globular clusters allow us to probe the star-formation history, chemical development, and mass distribution in their host galaxies. Of particular interest are globular clusters that reside in the remote haloes of their hosts, where dynamical timescales are long and these objects can offer additional insights into the history of interaction, merger, and accretion events in these systems. Galaxies in the Local Group (i.e., out to distances of roughly 1 megaparsec (Mpc)) are sufficiently close that they may be resolved into their constituent stars with present technology. Hence, these objects are attractive targets for detailed observational study, with the aim of inferring their evolutionary histories from present-day structure and

Figure 1.

A composite g,r, and i image of MGC1 from GMOS-North.

content. The “fossil record” contained in old and intermediate-aged stars in the Andromeda spiral galaxy (M31) is of special significance, as it is the closest large galaxy to our own and shares many characteristics in common with our Milky Way, suggesting a similar mode of assembly. Over the past decade, our group has been surveying M31 and its environs to unprecedented depths using widefield imagers attached to both medium and large telescopes. Our original survey used the Wide-field Camera on the 2.5-meter Isaac Newton Telescope to uncover a spectacular wealth of low-brightness substructure in M31’s inner halo (R < 30 kiloparsecs (kpc)), indicative of an active history of merger and accretion events in this galaxy. Our ongoing Pan-Andromeda Archaeological Survey (PAndAS) uses the Canada-France-Hawai’i Telescope (CFHT) MegaCam to extend this work to the far outer halo (R < 150 kpc). This has led to the discovery of many



fascinating new substructures, a gigantic stellar halo,

The fact that MGC1 exhibits a well-populated horizontal

and numerous new dwarf satellites. We have also been

branch that extends to quite blue colors is indicative of

using these survey datasets to search for new globular

a very old (> 10 billion years (Gyr)) and rather metal-

clusters in the remote regions of M31. To date, we have

poor stellar population, typical of many halo globular

discovered nearly 100 such objects, of which more than

clusters seen in the Milky Way. The MGC1 horizontal

60 lie at projected (on-sky) distances greater than 30

branch is evenly populated across the region of the

kpc from the galaxy’s center. Previously, only a handful

CMD harboring the instability strip, meaning that the

of such remote objects were known in M31, while the

cluster likely possesses a number of RR Lyrae stars.

Milky Way appears to possess just 10.

However, our imaging does not span a sufficiently long temporal baseline to search directly for the variability








expected from such members.

Spectrograph (GMOS) on the Gemini North telescope

Figure 2.

A color-magnitude diagram for MGC1 on the (g-i, i) plane, along with an example ridgeline (red line) from stellar evolution models calculated by the BaSTI (Bag of Stellar Tracks and Isochrones) group. The implied properties of the cluster are indicated.

to obtain high-resolution deep imaging of many of our new discoveries. The extremely high optical quality and stability of the GMOS instrument, in combination with the light-collecting ability of the 8-meter Gemini North mirror and the unrivalled atmospheric conditions that sometimes occur above Mauna Kea, provide image quality (IQ ) and depth in optical passbands that are unlikely to be bettered by other present-day terrestrial facilities. This combination is crucial for resolving our compact, distant cluster targets into individual stars and examining their internal contents.

The Most Isolated Globular Cluster in the Local Group One of the most intriguing newly discovered objects in

The shape and steepness of the red-giant branch allow

the M31 halo is a bright globular cluster named MGC1,

us to obtain estimates of the cluster’s metal abundance

which lies at a projected distance of 117 kpc from the

and line-of-sight distance, which we do by fitting

galaxy’s center. This already makes the cluster one of

ridgelines derived from both stellar evolution models

the most remote such objects known in the M31 system,

and from observations of various Milky Way globular

and indeed in the Local Group as a whole. We obtained

clusters. An example ridgeline from a set of stellar

deep imaging of MGC1 with GMOS-N on the night of

evolution models is shown over-plotted in Figure 2.

October 13, 2007, in the g, r, and i optical passbands.

Our fits demonstrate that MGC1 has a very low metal

The data were obtained under excellent conditions,

abundance, [Fe/H] = -2.3 +/- 0.2, which is comparable

resulting in an IQ of between 0.4 - 0.5 arcsecond across

to the lowest-metallicity globular clusters found in the

the three different filters. Our GMOS image of the

Milky Way and apparently substantially more metal-

center of the cluster is shown in Figure 1.

poor than the bulk of the field-halo stars observed in M31 at similarly large projected radii.

We obtained photometric measurements of all stars in


the image in each of the three passbands, allowing

In addition, our fits show that MGC1 lies much closer

us to construct color-magnitude diagrams (CMDs) for

to us, with a line-of-sight distance of approximately

the cluster. An example CMD, on the (g-i, i) plane, is

630 kpc, than do the central regions of M31, which

shown in Figure 2. The red-giant branch (hydrogen

fall at 780 kpc. Combined with its large projected



radius, this observation renders MGC1 the most

helium-burning-stars) of MGC1 are clearly visible; our

isolated known globular cluster in the Local Group

data reach at least one magnitude below the level of

by some considerable margin. It has a true (three-

the latter feature.

dimensional) distance of 200 +/- 20 kpc from the center





of M31. For comparison, the most remote Milky Way

MegaCam and Subaru/SuprimeCam images to extend

globular cluster is AM-1, which resides a ”mere” 120

the observed area around the cluster, we were able to

kpc from the galactic center. Our measurement shows

measure MGC1 stars to a projected distance of at least

that MGC1 lies in the extreme outer reaches of the

450 parsecs (pc) from the cluster center, and possibly as

M31 halo; its observed radial velocity (from a Keck

far as 900 pc (Figure 3). This renders MGC1 the most

telescope spectrum) is, nonetheless, within the probable

extended globular cluster hitherto studied (the Milky

M31 escape velocity, so there is no implication that

Way cluster NGC 2419 previously held this honor

MGC1 might be anything as exotic as an unbound or

with a tidal radius of roughly 200 pc). Interestingly,

intergalactic globular cluster. Instead, our measurements

the surface density of MGC1 falls off like a power-law

suggest that the globular cluster systems of large spiral

with radius as opposed to following a more typical

galaxies may be considerably more spatially extended

King-type profile, which would exhibit a sharp outer

than previously appreciated.

limit. This is broadly in line with expectations derived from numerical modelling of isolated globular clusters,

Our GMOS imaging revealed an additional remarkable

in which a core-halo structure is established over

feature of MGC1 – its spatial extent. We traced the

several billion years, and suggests that MGC1 has been

cluster-centric radial distribution of stars in our images,

evolving in isolation in the outer halo of M31 for a very

and to our surprise found that we could identify

considerable period of time.

cluster members out to the very edge of the GMOS field. Indeed, by using archival wide-field CFHT/

The Nature of Extended Clusters Another particularly intriguing subset of the remote objects we have discovered in the M31 halo is the class of so-called “extended clusters.” As shown in Figure 4 (next page), these systems occupy an unusual region of size-luminosity space, encroaching on the traditional gap between classical globular clusters (which are compact and do not appear to contain significant amounts of dark matter) and dwarf spheroidal galaxies (which are diffuse and dominated by dark matter). It is unclear whether extended clusters are more akin to classical globulars or dwarf spheroidals, or are some kind of transition population. To add to the puzzle, the Milky Way apparently does not possess similarly extended luminous clusters as those observed in M31, but does harbor a few diffuse low-luminosity globular clusters. The Milky Way also contains a number of “ultra-faint” dwarf galaxies which have roughly comparable sizes to the M31 extended clusters but exhibit tiny luminosities (note, M31 may also possess such objects; however, they would be too faint to be detected in presently available survey data). Unveiling the nature of extended clusters is of obvious importance to help us better understand the faint end

Figure 3.

Radial g-band surface brightness profile for MGC1, derived from our GMOS imaging (solid black circles and magenta triangles) and, at large radii, from archival CFHT/ MegaCam and Subaru/SuprimeCam data (red stars and open blue circles). The horizontal dotted line marks the level of background contamination, while the vertical dashed line marks the maximum reliable resolution of the images. The commonly used King family of models does not describe the data well (top panel); instead the profile breaks to a powerlaw fall-off in its outer regions (bottom panel).

of the galaxy luminosity function. Following on from our successful observations of MGC1, we are leading an international Gemini observing program (24 hours spread across allocations from the UK, Canada, and Australia) aimed at addressing this problem. Our targets consist



Figure 4.

This work has been done in collaboration with a

Various stellar systems on the sizeluminosity plane. The extended clusters we have discovered in M31 apparently occupy an intriguing region, encroaching on the traditional gap between globular clusters on the left and dwarf spheroidal galaxies on the right, and of similar size to (but much brighter than) several of the new “ultra-faint” dwarf galaxies recently discovered in the Milky Way halo (labelled).

number of additional people, including Mike Irwin (University of Cambridge), Nicolas Martin (MPIA, Heidelberg), Nial Tanvir (University of Leicester), Alan McConnachie (HIA, Victoria), Rodrigo Ibata (Université de Strasbourg), Scott Chapman (University of Cambridge), and Geraint Lewis (University of Sydney). The work concerning MGC1 is in press at the Monthly Notices of the Royal Astronomical Society (2009). For more information see: Ferguson A., et al., 2002, AJ, 124, 1452 Huxor A., et al., 2005, MNRAS, 360, 1007 Huxor A., et al., 2008, MNRAS, 385, 1989 Ibata R., et al., 2007, ApJ, 671, 1591 Mackey A., et al., 2006, ApJ, 653, L105 Martin N., et al., 2006, MNRAS, 371, 1983

of a representative sample of 11 extended clusters in the M31 halo. They populate the full range of observed sizes, luminosities, and colors. We obtained deep GMOS-N imaging of these objects in the g and i optical passbands between July 6, and October 31, 2008. We are still in the process of analyzing these observations; however, some preliminary images and CMDs are shown in Figure 5 to highlight the exquisite quality of the data (some of our images have IQ approaching 0.3 arcsecond). Ultimately, our imaging will allow us to more accurately constrain the region of size-luminosity space occupied by the M31 extended clusters, as well as providing detailed information about the constituent stellar populations of these peculiar objects.

Figure 5.

Deep GMOS-N i-band images of three of our newly discovered M31 extended clusters, along with preliminary (g-i, i) CMDs that show these objects are very likely ancient, metal-poor systems.



Dougal Mackey is a Marie Curie Excellence postdoctoral fellow at the University of Edinburgh Institute for Astronomy (IfA). He can be contacted at: Annette Ferguson leads a Marie Curie Excellence Team at the University of Edinburgh IfA, where she also holds a Readership in Astronomy. She can be contacted at: Avon Huxor is a Leverhulme Fellow in astrophysics at the University of Bristol. He can be contacted at:

Near-IR Science by Mariska Kriek & Pieter van Dokkum

The â&#x20AC;&#x153;Wild Youthâ&#x20AC;? of Massive Galaxies Owing to the finite speed of light and state-of-the-art instrumentation on large telescopes, astronomers are able to identify and study galaxies throughout the history of the universe and witness their evolution over cosmic time. During the first several billion years after the Big Bang, the universe was essentially a large star-forming factory, with gas being converted into stars at ferocious rates. Nonetheless, as our team showed several years ago using Gemini, mature galaxies can already be found some 3 billion years after the Big Bang, when the universe was only about 20% of its current age. These galaxies resemble nearby massive elliptical galaxies in terms of their red colors, very low star-formation rates, and their large stellar masses. However, appearances are deceiving. Just a couple of years ago, we were living in ignorance of what might currently be one of the largest puzzles in galaxy evolution. Images taken with the Hubble Space Telescope in 2008 showed that the structure of these red massive galaxies in the young universe differs considerably from the red elliptical galaxies of similar mass in the nearby universe: their sizes, parameterized by the radius that contains 50% of the light, are about a factor of five smaller. While the half-light radii of todayâ&#x20AC;&#x2122;s massive elliptical galaxies are about 5 kiloparsecs (kpc), the red galaxies typically have half-light radii of 1 kpc! Since the average stellar density of a galaxy is proportional to the cube of the inverse of the radius, these distant red galaxies were extremely dense systems. Moreover, as they were already as massive as the most massive galaxies in the nearby universe, they would have to increase in size tremendously without growing much in mass. The difficulties associated with the mass and size determinations of such distant galaxies caused skepticism among members of the astronomical community. How well can we actually measure these fundamental



Figure 1.

(a) Spectrum of 12550, obtained over the course of 6 nights with GNIRS on Gemini South. (b-d) Hubble Space Telescope image of the object, along with a model of the Hubble image and the difference between the image and the model are shown at right.

properties? Were we underestimating the sizes, or

old. The observations were extremely challenging,

perhaps overestimating the stellar masses? Due to the

as the galaxy is very faint, owing to its large

large cosmological distances, the observed galaxies

distance. Furthermore, due to the expansion of the

are faint, and there were worries that fainter starlight

universe, the emitted optical light of the galaxy has

might have been missed. Also, the technique used

been shifted to longer wavelengths. Thus, we had

to determine the masses raised some eyebrows. At

to take observations at near-infrared wavelengths,

these distances, masses are commonly derived using

which are greatly affected by Earth’s atmosphere.

an estimate of the mass-to-light ratio, which is based

We observed the galaxy during six full nights under

on the colors of the galaxies. Together with the

perfect weather conditions, for a total of almost 30

observed luminosity of the galaxy, a “rough’’ distance

hours of observations. The final result is probably

determination, and many more assumptions, this

the deepest spectrum at near-infrared wavelengths

yields a stellar mass estimate. Taken altogether, these

ever taken.

assumptions easily lead to uncertainties of a factor of two. The ultimate test to verify the high stellar

This ultra-deep spectrum allowed us to estimate the

densities of the distant red galaxies is to measure

typical velocities of stars by measuring the Doppler

the velocities of their stars. These velocities are a

broadening of stellar absorption lines. The stars in

sensitive probe of density, as they are proportional

the galaxy have different motions relative to each

to the square root of the mass divided by the radius.

other, so the spectra of the individual stars are

Therefore, in massive galaxies with small sizes,

shifted slightly with regard to one another. The final

the stars should be whizzing around at very high

effect is that the absorption line of individual stars


will be smoothed out into a broader feature when combining all the light. The amount of broadening






tells us the typical relative motions of the stars.

(GNIRS) we succeeded (for the first time) in


measuring the stellar velocities in a galaxy (1255-0)

Based on our initial measurements of the stellar mass

seen when the universe was less than 3 billion years

and size, and on Newtonian physics, we expected


Figure 2.

that was used for this work can only observe one galaxy at a time, although with full near-infrared wavelength coverage. With the multi-object nearinfrared spectrograph FLAMINGOS-2, (currently in commissioning) these studies will be much more efficient, as more galaxies can be observed at the same time. Also, the quality of the spectrum presented here is far from what could theoretically be achieved with 8-meter-class telescopes. There is potential for considerable improvement in the capabilities of single-object near-infrared spectrographs, and a successor to GNIRS with higher resolution and much greater sensitivity would have an enormous

The relationship between stellar mass, dynamical mass, radius re and velocity dispersion s for nearby galaxies (black points) is shown. The location of 1255-0 is indicated by the red point. For its mass, the galaxy has a very small radius and a very high velocity dispersion, as compared to nearby galaxies.

impact on studies of distant galaxies.

to find velocities of over 500 kilometers per second (km/s) for this galaxy. The velocity dispersion that we determined from the Gemini spectrum is in excellent agreement with our expectations: 510 km/s, with an uncertainty of about 100 km/s. This confirms our initial result, namely that the stellar density in this galaxy was extremely high 11 billion years ago. Thus, the small sizes and high masses appear to be correct, and cannot easily be attributed to measurement errors. Our new result confirms the existence of a major puzzle that still remains to be solved: how is it possible that these galaxies increase so much in size

For more information see: van Dokkum, P. G., Kriek, M., and Franx, M. 2009, Nature, 460, 717 Kriek, M., van Dokkum, P. G., Labbe, I., Franx, M., Illingworth, G. D., Marchesini, D., and Quadri, R. F. 2009, ApJ, 700, 221 Mariska Kriek is a post-doctoral fellow at the department of Astrophysical Sciences, Princeton University. She can be reached at: Pieter van Dokkum is a professor at the Astronomy Department, Yale University. He can be reached at:

and decrease in stellar density over time, without becoming much more massive? There are several theories proposed to explain how distant compact galaxies may have been puffed up in the past 11 billion years. It is plausible that they are the centers of nearby massive elliptical galaxies, and that the outskirts were built up later by mergers with other smaller galaxies. This would change the energy distribution in the galaxy and could lower the velocities of the stars while building up low-density wings. Other processes have been proposed as well. For example, a quasar phase might be responsible for expelling gas from the galaxies, making them larger and less dense. Whereas the responsible process still remains to be identified, it is evident that the compact galaxies in the early universe do not live a quiet and undisturbed life. Clearly, more studies are needed, and exciting times await. The Gemini Near-Infrared Spectrograph



Spectroscopic Methods by Adam Muzzin & Howard Yee

Spectroscopy of High-Z Galaxy Clusters with Band-Shuffle Rich clusters of galaxies are some of the most spectacular objects in the universe. They are the largest gravitationally bound structures, containing hundreds of massive galaxies and thousands of smaller ones all whizzing around at thousands of kilometers per second within a relatively compact 10 to 20 cubic megaparsecs. Staring at Hubble Space Telescope images of nearby clusters is a constant reminder of just how small we are compared to the rest of the universe. It is well known that the galaxies in clusters are different from those that aren’t in clusters–which are typically referred to as “field” galaxies. Clusters contain massive galaxies that are non-star-forming, usually elliptical or type S0, whereas field galaxies are often less massive, spiral-shaped, and have star-forming regions. For decades, astronomers have been trying to understand why the cluster population is so different from the field population, but no consensus has emerged yet. We know the differences are probably related to the cluster environment, which contains hot x-ray-emitting gas, experiences high-speed encounters between galaxies, and shows extreme tidal effects from the huge dark-matter halo. Yet, we are uncertain about exactly how these factors alter the cluster galaxy population. An obvious way to get a better handle on the problem is to observe the most distant clusters, meaning those at z > 1, to see how their galaxies differ from nearby clusters, as well as their counterparts in the field at z > 1. Galaxies are more active at higher redshifts, which should further highlight the contrast between cluster and field populations. But, before high-redshift clusters can be observed, we have to find them. It’s hard work. The intra-cluster gas in distant clusters is faint in the x-ray regime because of surface-brightness dimming that scales as Lx ~ (1 + z)-4. Cluster galaxies are also faint, and Doppler-shifting means most of their light is not received at optical



wavelengths, but in the infrared. As a result, deep x-ray or infrared surveys are needed to detect distant

Spectroscopic Confirmation of SpARCS Clusters With Band-Shuffle on Gemini

clusters—both of which are still hard to come by these days.

To prove that SpARCS is selecting real clusters, we wanted to confirm the redshifts and masses of some of

In addition to the need for deep observational data,

our candidate clusters with spectroscopy. However,

rich clusters of galaxies are rare objects. Current

obtaining redshifts and velocity dispersions for z >

lambda cold dark matter (λCDM) halo models

1 clusters is observationally intensive work. Even at

predict about one Coma-mass cluster (M ~ 1015 MSun) at z > 1 in every 50-100 square degrees of sky. Even

z > 1, where most galaxies are forming stars, cluster

clusters half that massive only occur once in every

requiring long exposures to obtain high signal-

2 - 4 square degrees of sky surveyed. Putting it all

to-noise on the continuum to detect absorption

together, we need to have both deep- and wide-field

features. Furthermore, their extreme distances mean

surveys to find distant clusters.

that even massive clusters are small in angular size,

galaxies are still frequently devoid of emission lines,

with most of their galaxies appearing in an area of

The SPARCS Survey

sky that is only about 1-2 arcminutes across.

Given the observational challenges, it should come

As an example, the left panel of Figure 1 (next page)

as little surprise that the number of confirmed

shows a color image of a massive z = 1.157 cluster

massive clusters at z > 1 is still quite small: around

in a 6 × 6 arcminutes field of view (f.o.v.)—roughly

10-30, depending on the definitions of “confirmed”

the f.o.v. of the Gemini Multi-Object Spectrograph

and “massive.” In 2005, the public release of the

(GMOS) and other multi-object spectrographs.

Spitzer Wide-area InfraRed Extragalactic legacy

Canonical multi-object spectroscopy (MOS) requires

survey (SWIRE) from the Spitzer Space Telescope presented a golden opportunity to look for more

long slits, ~ 10 arcseconds long, in order to facilitate the subtraction of sky lines, which means even in

clusters using infrared/optical methods. At 50 square

ideal cases a typical mask can only get 6-12 slits

degrees, SWIRE occupies a unique parameter

across the central region of distant clusters, a pretty

space. It is a wide enough field to find a significant

dismal return of cluster galaxy spectra per mask.

population of massive z > 1 clusters, and, due to Spitzer’s power in the infrared, it also reaches deep

The nod-and-shuffle (N&S) mode available on

enough to reliably select z > 1 candidate clusters.

GMOS offers a major improvement in slit placement over regular MOS modes. Because targets are nodded

To take advantage of the large SWIRE area, we

across the slit for sky subtraction, the required length

decided to use the well-proven Cluster Red-Sequence

of the slit is only two to three times the angular

(CRS) method that looks for clusters as over-

size of the object. For distant galaxies, typically ~ 1 arcsecond across, it means only requiring slits ~ 3 arcseconds across. This provides a major gain over

densities of galaxies in a combined color-position space. This method has been used extensively to look for clusters up to z ~ 1 in the 100-square-degree RCS-1 survey and the next generation 1000-square-

the typical 10 arcseconds long slits. There are two

degree RCS-2 survey. We conducted a deep survey

be performed: “micro-shuffle” and “band-shuffle”.

of SWIRE in the z-band in order to use a combined

The majority of N&S spectroscopy with GMOS

z - 3.6-micron color to select clusters at z > 1. This

(such as the Gemini Deep Deep Survey (GDDS))

project, the Spitzer Adaptation of the Red-sequence

has been done in the micro-shuffle mode, where

Clusters Survey (SpARCS), was the first to try to employ the CRS method to select z > 1 clusters using infrared data; it remains the largest such survey for z > 1 clusters to date. The survey contains several hundred candidate clusters at z > 1, with estimated masses between ~ 5 × 10 MSun and 1 × 10 MSun. 13


observing modes in which N&S observations can

shuffled charges are stored directly above or below the location of the slit. This means a 3-arcseconds slit requires a blank space of 3 arcseconds above or below it to store shuffled charge, increasing the effective slit area to 6 arcseconds. This is still better than conventional spectroscopy, but is not ideal for clusters.



Figure 1.

Left Panel: A g,z, 3.6μm color image of the cluster SpARCS J161641+554513 at z = 1.157. The field of view (f.o.v.) of the image is 6 × 6 arcminutes, approximately the f.o.v. of GMOS. The yellow boxed area shows the region that can be used for “band-shuffle” observations. Middle Panel: Spectra within the band-shuffle region, from a spectroscopic mask observed with LRIS on Keck. Right Panel: Spectra from a band-shuffle mask on GMOS. The 3× smaller slit length from N&S, combined with a band-limiting filter, allowed 52 slits to be fit within the region, making GMOS about 4 times more efficient than LRIS for obtaining spectra of cluster galaxies at z ~ 1.

Figure 2.

Left Panels: g,z, 3.6μm color images of three z > 1 clusters from the SpARCS survey with redshifts and masses confirmed by GMOS observations. Right Panels: Same as left panels, with white squares representing confirmed cluster members, and green circles representing confirmed foreground and background galaxies. These clusters are the most distant discovered with the CRS method to date and are some of the most distant galaxy clusters known.


In band-shuffle mode, the entire central region of

far the most distant cluster detected with the CRS

the chip is shuffled to the top or bottom of the

method, demonstrating that the highly efficient two-

chip, which means no space is required between

filter CRS method works extremely well for finding

slits. While it is technically less efficient in the

massive clusters, even out to the highest redshifts.

total use of the GMOS chips (only one third of the area is used, whereas up to half can be used with micro-shuffle), it is perfect for high-z clusters. With

Next Step: Gemini Cluster Astrophysics Spectroscopic Survey (GCLASS)

a band-limiting filter we can now typically place 35 - 55 slits within an area of about 3.0 × 1.7 arcminutes

After our initial observing runs with GMOS, we

around these clusters.

quickly realized how powerful the instrument would be for a detailed follow-up study of distant clusters.

In the middle panel

Individual z > 1 x-ray

of Figure 1 we show

and optic al/infrared

a  spectroscopic


selected clusters have



been studied; however,

c o nv e n t i o n a l  M O S

no survey of a large of


z > 1

with LRIS on Keck.


The right panel shows

c l u s t e r s  h a s  b e e n



done. The size of the

band-shuf f le masks.

SpARCS survey meant

Note the considerable

that we had numerous



rich z > 1 clusters that

over LRIS in packing

could be followed up

in the slits using band-

in detail.




shuffle mode. As of semester 2009B, GMOS has been

we are now halfway



confirming redshifts



through observations



for the Gemini





of the first z > 1 cluster


c a n d i d a t e s  i n  t h e

(GCLASS). This large,




200-hour project aims


to obtain spectra of

images of three SpARCS

50 galaxies in 10 rich

clusters, all confirmed

clusters at 0.87 < z < 1.34


survey. 2


with GMOS. SpARCS J003550-431224 at z = 1.335,

selected from the SpARCS survey, making it the

in particular, is an important find. It is currently

state-of-the-art distant cluster project. The scope of

one of the most distant known clusters and is by

the science covered by the project is large. Here are


Figure 3.

For more information see: Muzzin, A., et al. 2009, ApJ, 698, 1934 Wilson, G., et al. 2009, ApJ, 698, 1943 Eisenhardt, P., et al. 2008, ApJ, 684, 905 Stanford, S. A., et al. 2006, ApJ, 646, L13 Muzzin, A., Wilson, G., Lacy, M., Yee, H.K.C., & Stanford, S. A. 2008, ApJ, 686, 966 Gladders, M. D. & Yee, H.K.C. 2005, ApJS, 157, 1 Gladders, M. D. & Yee, H.K.C. 2000, AJ, 120, 2148

some highlights: the first measurement of accurate dynamical masses for a sample of z ~ 1 clusters, which will be a crucial factor in studying dark energy with cluster abundances; the first measurement of the average mass and light profile of a massive dark

Adam Muzzin is a postdoctoral associate at Yale University. He can be reached at: Howard Yee is professor of astrophysics at the University of Toronto. He can be reached at:

matter halo at z ~ 1, from “stacking” the velocity data on the clusters; and a extensive look at the stellar

Rest-frame stacked spectra of the average galaxy in six out of ten clusters in the GCLASS project. Prominent spectral features such as [OII] emission and hydrogen Balmer absorption are (Hδ, Hζ, Hη, Hθ) marked. The clusters show a diverse range in [OII] emissionline strength and Balmer absorption, which clearly shows the importance of a large sample of clusters in understanding their stellar populations.

populations of galaxies in a sample of homogenouslyselected rich clusters at z ~ 1. In Figure 3 we plot stacked spectra of cluster members from the first ~40 percent of observations from the GCLASS project. With only a few of the masks on about half of the clusters, a significant diversity in both the strength of the [OII] line (typically from star formation) and the Hδ line (typically an indicator of a “post starburst” phase) can already be seen. Figure 3 not only demonstrates the exquisite quality of spectra that can be obtained for distant galaxies using N&S, but also shows the cluster-tocluster variations in the stellar populations of their galaxies. From Figure 3, it is already clear that to get a fair sample of galaxies in distant clusters, a sample of 10 or even more clusters is crucial. The







spectroscopy on these clusters, and I look forward to summarizing the results from the GCLASS survey in a future article. Observations for GCLASS are expected to conclude in semester 2010B. The work presented in this article has been done in close collaboration with Gillian Wilson (University of California, Riverside) who serves as PI of the SpARCS/GCLASS projects. The SpARCS/GCLASS team consists of 26 members from 19 institutions.



AO Innovations by Richard McDermid, Davor Krajnović & Michele Cappellari

Black Holes & Revelations Using Open-loop Adaptive Optics Supermassive black holes are thought to reside at the heart of many, if not all, galaxies in the nearby universe. These weighty beasts can account for as much as 2% of the entire mass of a galaxy, all containing up to several billion solar masses in a compact central region smaller than our solar system. Although the actual “size” of a supermassive black hole is small, it is thought to play an integral role in the formation and evolution of its host galaxy, primarily by accreting material and generating powerful jets of energy. Most black holes in the local universe are no longer in this accreting phase but their gravitational effects on nearby stars can be detected. Measuring the orbital velocities of stars or gas within the radius of influence— where the gravitational pull of the black hole dominates over the effect of the host galaxy—can reveal the gravitational signature of the unseen supermassive black hole. More importantly, by accurately modeling these motions, it is possible to derive the most fundamental property of the black hole: its mass. When galaxies merge, their respective central black holes are thought to fall to the center of the remnant galaxy, forming a black hole binary system. The binary can eject stars that pass close to the galaxy center and by doing so, loose angular momentum and eventually coalesce. It is thought that this ejection of stars from the central region is a mechanism for forming the low-density galaxy cores found in massive early-type galaxies. This so-called “core-scouring” mechanism leaves a specific imprint on the orbital structure of the central region, since stars on radial orbits are preferentially ejected, leaving a higher fraction of stars on more circular (or “tangential”) orbits.



To understand the orbital structure around the black

realizing spatial resolutions close the theoretical

hole, and hence gain an insight to the formation

limit of the telescope. This effectively gives Gemini

mechanism of the galaxy, the main issue is to

a spatial resolving power comparable to the Hubble

spatially resolve the kinematics close to the black

Space Telescope (HST), whilst having an order of

hole’s radius of influence. For nearby galaxies, this is

magnitude larger light-collecting area and a broader

typically only a few tenths of an arcsecond on the

range of instrumental capabilities.

sky. At these scales, the Earth’s atmosphere blurs the information, mixing up light from the galaxy

Applying all this new technology to the topic of

nucleus with light from further out in the galaxy,

supermassive black holes has not been entirely

and greatly diluting the signature of the black hole.

straightforward. The main issue is that, although LGS-AO allows fainter guide sources to be used,

Natural guide star adaptive optics (NGS-AO) on

galaxies that may have a resolvable radius of influence

large telescopes can help overcome the limitations

tend to be nearby and spatially extended on the sky.

imposed by Earth’s atmosphere. However, this

This means that guide stars are either superimposed

technique requires bright guide stars that appear

on the galaxy’s light (requiring them to be relatively

close to the nearby galaxies of interest, which are

bright to be useful) or they are positioned so far

rare. Laser guide star adaptive optics (LGS-AO)

from the galaxy nucleus that the spatial resolution

allows much fainter guide sources to be used, greatly

is degraded due to anisoplanatism—incoherence of

increasing the fraction of the night sky available

the atmosphere between two different positions on

for AO-assisted observations. LGS-AO creates an

the sky.

artificial star, or reference source with which to measure the atmospheric distortion by making use of

To get around this problem, we have used a new

a one-kilometer-thick layer of the Earth’s atmosphere

technique at Gemini North known as the “open-

composed largely of sodium, located at an altitude

loop” focus model. While the nuclei of most nearby

of around 90 kilometers. A powerful laser beam that

galaxies are too faint and extended to measure the

has a wavelength identical to a resonant wavelength

telescope focus reliably, they can still be used to

of the sodium atom (which turns out to be the

determine tip/tilt. Information on the telescope

same color as most common street lighting) is fired

focus is required to account for possible changes in

into this layer. The laser light stimulates a narrow

the distance from the telescope to the sodium layer,

column of gas within this sodium layer, causing the

which can drive the telescope out of focus.

gas to emit light. With this technique, instead of trying to measure As seen from the telescope, a bright point of light

the focus dynamically on the galaxy nucleus, the

is created high in the atmosphere, allowing the

assumed distance to the sodium layer (and therefore

atmospheric distortion to be measured and corrected

the telescope focus) is controlled by an open-loop

by the AO system. A natural reference source is also

model. This is a geometric model of how the distance

needed to measure the bulk motion, or “tip/tilt,”

to the sodium layer changes as the telescope tracks

caused by the atmosphere, as well as to monitor

a target on the sky and requires no real-time focus

relative changes in focus caused by a varying

measurements. The telescope and AO system are

sodium layer altitude. This natural source, however,

first “tuned,” or optimized, on a nearby star, which

can be several magnitudes fainter than is needed for

calibrates the altitude of the sodium layer. This

correction of higher-order modes, therefore giving a

altitude is then kept fixed for the duration of the

much larger sky-coverage than NGS-AO.

science observation. The galaxy is acquired, and the nucleus is used to measure tip/tilt, while the higher-

The advent of LGS-AO opens a new era of high spatial

order atmospheric distortions are measured from

resolution observations from the ground. Moreover,

the laser beacon. The open-loop model controls the

the development of integral-field spectroscopy and

changes in focus as the distance to the sodium layer

infrared capabilities allow a two-dimensional region


to be spectrally mapped simultaneously, while



Figure 1.

NIFS stellar kinematic maps for NGC 2549 (top) and NGC 524 (bottom). The velocity profile is parameterized by the standard Gauss-Hermite series, giving the mean velocity, V, the velocity dispersion, σ, and higher-order moments h3 to h6. Note the observed rotation, the central increase in σ, and the anti-symmetric signatures in the higher moments. The data are spatially binned with a threshold signal-to-noise ratio of 50.

We used this technique together with the Near-

To estimate the degradation of the point-spread



function (PSF) due to the use of the open-loop

measure the stellar kinematics around the nucleus of




model, we took NIFS observations of the star used

two nearby early-type galaxies. Figure 1 shows the

to tune the AO system both before the galaxy was

measured stellar kinematic maps for the two galaxies,

observed (where all loops were optimized) and after

NGC 524 and NGC 2549, which were observed

approximately one hour of galaxy observations (with

for 3.5 and 2.5 hours, respectively, on-source (i.e.

the focus loop left open, as for the galaxy). Figure 2

excluding blank sky fields). Neighboring spectra have

compares the PSFs derived from these observations,

been binned together in the outer regions, where the

which show negligible degradation due to changes in

galaxy is fainter to ensure a minimum signal-to-noise

the sodium layer or errors in the open-loop model.

ratio of around 50 in each spectrum. The central

This is only one instance, of course, and the stability

spectra have a higher signal and are unbinned, thus

of the sodium layer altitude depends on the weather

preserving the spatial information delivered by the

conditions for a given night. However, it seems that

LGS-AO. This excellent data quality allows the

this mode gives a reliable performance, at least over

higher-order moments of the velocity distribution to

timescales of around an hour.

be measured, using a 6th-order Gauss-Hermite series.

Figure 2.

Comparison of the NIFS point-spread function (PSF) derived from a star close to NGC 524 used to tune the telescope and AO system. The observations were made before (left) and after (right) observing the galaxy. The “after” images were taken without re-calibrating the sodium layer altitude. This indicates any degradation due to using the open-loop focus model. The parameters of the double-Guassian fit used to parameterise the NIFS PSF are almost identical, showing that no significant degradation occurred.


The higher-order moments are essential to derive

To infer the mass of the black hole, and the stellar

the detailed orbital make-up around the black hole.

orbits around it, requires a dynamical model of the

We combine the NIFS data with wide-field integral

galaxy. The mass and distribution of the galaxy’s

field kinematics of comparable quality, but lower

stars are modeled using imaging information (from

spatial resolution, from the SAURON spectrograph

HST and ground-based telescopes), converted to

at the William Herschel Telescope.

mass via the mass-to-light ratio: a free parameter


Figure 3.

Comparison of the observed NIFS velocity dispersion (left column) and the best-fitting model (second to left column) for NGC 2549 (top) and NGC 524 (bottom). Models with too small and too large a black hole (second from right and right columns respectively) are also shown for comparison. Note how only the central pixels differ significantly. The NIFS data have been symmeterized for this comparison.

in the dynamical model. In addition to this, a

Figure 3 shows a comparison of the central velocity

central dark mass is included, which represents the

dispersion (which is generally the most sensitive

black hole, and gives another free parameter. Lastly,

kinematic tracer of the black hole) from NIFS with

the angle at which the galaxy is viewed must be

that of the fit from the best-fitting model, as well as

assumed, since the shape of the galaxy we see is a

models with smaller and larger black holes. While

two-dimensional projection on the sky of a three-

most of the field is unchanged between the different

dimensional body. The objects we are modeling are

models, the very central regions of the NIFS data

well described by a galaxy symmetrical about its axis

show variations. It is these subtle differences that

of rotation, while the viewing direction is uniquely

indicate the presence and mass of the central black

described by the inclination of the system. For both


galaxies, good estimates of this inclination angle are available from the imaging data, so we assume these

Thanks to the high-quality NIFS integral field data

fixed values. Dark matter is included in the model

,coupled with our general dynamical modeling,

implicitly through the mass-to-light ratio, making the

we can also infer the orbital structure close to the

assumption that the dark matter is distributed in the

black hole. Figure 4 shows a measure of the balance

same way as the starlight. In practice, dark matter

between radial and tangential orbits each galaxy.

is not expected to dominate the stellar dynamics in

NGC 524 shows a very distinct change inside the

the central regions of these galaxies.

black hole radius of influence towards tangential anisotropy, meaning a relatively higher fraction of


Figure 4.

Radial profiles showing the balance between radial and tangential orbits for NGC 524 (left) and NGC 2549 (right). Colored lines indicate different angles from the equatorial plane, increasing from red to blue. Dot-dash lines indicate the width of the PSF core, and the dashed lines show the radius of influence. NGC 524 shows a strong tangential anisotropy within the radius of influence. NGC 2549 shows a weaker signature, but the radius of influence is not as well resolved.


orbits going around the galaxy center than passing

This work is presented in full in KrajnoviÄ&#x2021;,

through it. This is compelling evidence that the

McDermid, Cappellari, and Davies, MNRAS, in

central core region of this galaxy was formed by

press. We gratefully acknowledge Chad Trujillo

a black hole binary following a galaxy merger that

and the staff at Gemini North for developing the

ejected the central stars on radial orbits, leaving

observing techniques using the open-loop focus

a depleted core region biased to tangential orbits.

model method for our observations. We note that

The picture for NGC 2549, which does not have a

this observing mode is generally offered at Gemini

low-density central core, is less clear. There is only

North as of semester 2009A.

weak evidence for any change in anisotropy close to the black hole, suggestive of a different formation

Richard McDermid is a Gemini science fellow at Gemini

history to NGC 524.

North. He can be reached at:

Using the open-loop focus model technique, LGS-AO

Davor KrajnoviÄ&#x2021; has recently moved from the University of

at Gemini has allowed unique, high-quality spectral

Oxford to ESO as a fellow. He can be reached at:

maps of these galaxy nuclei to be made at very high

spatial resolution. These observations have revealed the presence of a supermassive black hole residing

Michele Cappellari is an STFC advanced fellow at the

at the center of each galaxy. More importantly,

University of Oxford. He can be reached at:

they provide evidence that early-type galaxies

with different central light characteristics (cusps and cores) have correspondingly different orbital structures, indicative of distinct formation scenarios. This adds new insight into the connections between supermassive black holes, galaxy nuclei, and the evolution of early-type galaxies.



by Nancy A. Levenson

Recent Science Highlights A Surprise Collision on Jupiter Australian amateur astronomer Anthony Wesley discovered a new feature on Jupiter on July 19, 2009. This new spot, which appeared dark at optical wavelengths, was located near Jupiter’s south polar region and was likely the result of a comet or asteroid impact. Scaling from observations of Comet D/1993 F2 (Shoemaker-Levy 9), which ran into Jupiter almost exactly 15 years ago, an impactor only a few hundreds of meters in diameter could have caused the new spot.

Figure 1.

Gemini observed the consequences of the impact at mid-infrared wavelengths, using both MICHELLE on Gemini North and T-ReCS on Gemini South. Some of the measurable effects include thermal radiation following heating of the atmosphere, so the impact site was bright in the mid-infrared. Signatures of particular constituents such as methane, ammonia, and aerosols, are also evident in the data. Later observations of the impact site tracked the dispersal of the debris feature and are providing information about Jupiter’s atmosphere and its winds. Heidi Hammel (Space Science Institute), with colleagues Imke de Pater (University of California, Berkeley), and Glenn Orton and Leigh Fletcher (both of Jet Propulsion Laboratory), obtained narrow-band images with MICHELLE through seven filters, with the help of Gemini staff: Tom Geballe, Rachel Mason, Chad Trujillo, Paul Hirst, and Tony Matulonis. Two of those images (at 8.7 and 9.7 μm) provide a spectacular result (Figure 1). The impact was truly unexpected. Gemini’s multi-instrument queue enabled data collection within 24 hours from when the team proposed the observations.


A mid-infrared false color image of Jupiter, from MICHELLE observations at 8.7 (red) and 9.7 microns (blue). The new impact site appears as a bright region at the center of the bottom of Jupiter’s disk. The size of this feature is consistent with an impacting object a few hundreds of meters in diameter.


High-Pressure Outflow from M82

Figure 2.

The central region of M82 at optical wavelengths (HST/ ACS image) showing the five 7- by 5-arcseconds fields observed with the GMOS-North IFU (white and black rectangles). Dashed lines mark some of the regions of most intense star formation.

resulted in two papers in the Astrophysical Journal, and the relationship between the inner and outer wind

The nearby starburst galaxy M82 is an archetype of its

regions is the planned subject of future work.

class. Individual regions of star formation in the galaxy can be measured accurately to account for the dynamics

The team’s GMOS data show that the rotation axes of

of intense star formation and the resulting outflowing

the gas and stars are offset, which provides supporting evidence that interaction with the nearby galaxy M81 triggered the starburst. While bright, spectrally narrow emission is common throughout the data, a diffuse broad component is also present, which the authors associate with turbulence on the surfaces of clouds in the dynamic environment. One region in particular shows evidence for an outflow channel for hot gas, which entrains the cooler, denser material. If the material eventually escapes the galaxy’s potential well, such outflows would be a mechanism for the chemical enrichment of the surrounding intergalactic medium. However, the team finds few similar coherent outflows in the observations, which may be a consequence of the irregularity of the ages, masses, and locations of the stellar clusters that drive

winds driven by the collective effect of stellar winds

them. In general, the electron density of the ionized

and supernovae. An international team led by Mark

interstellar medium decreases along the galaxy’s minor

Westmoquette (University College, London) used the

axis (Figure 3), and the similarity of the density with

integral field unit (IFU) of the Gemini Multi-Object

spectral width and radial velocity in a second off-axis

Spectrograph (GMOS) to study the ionized gas and

region could be a sign of another outflow channel.

stellar dynamics at M82’s core (Figure 2). This work

Figure 3.

Electron density measured from [SII] line ratios in the GMOS IFU data. The density generally declines along the minor axis of M82, and the highest densities arise in very compact regions, though not in the starburst complexes themselves.



Dynamics of an Ultracompact HII Region

Figure 4.

could account for the observed image positions). Structure in the lensing galaxy’s halo is one possible

The central massive stellar source of the ultracompact

origin for a more complex mass distribution. If dark

HII region K3-50A, located within the Milky Way

matter forms the substructure, lens models could

Galaxy at an estimated distance of 7 kiloparsecs, remains

provide important constraints on its nature and role in

hidden at near-infrared wavelengths, buried in its dusty

galaxy formation, specifically identifying some of the

native environment. However, new observations reveal

low-mass satellite galaxies that cold dark matter models

interesting dynamics and structure within the 0.1 parsec

predict in abundance. University of Washington graduate student Chelsea MacLeod and collaborators found a much more mundane explanation for the flux anomalies: a second galaxy. They obtained images of the system at 11 microns with MICHELLE on Gemini North. Because the quasar is extended in the mid-infrared, the images are free from the complication of microlensing by stars, and they suffer less from extinction. The combination of the main lens and the nearby galaxy reproduce

Velocity for the principal component of the Brγ line, measured in kilometers per second, with the mean velocity subtracted to show spatial variations. The bipolar structure is not aligned with any known point source or largescale ionized flow detected at radio wavelengths. Red filled circles and yellow diamonds mark previouslydetected sources at near-infrared and mid-infrared wavelengths.

both the positions and flux ratios of the quasar images (Figure 5).

Figure 5.

scale of the ionized gas. Robert Blum (National Optical Astronomy Observatory, Gemini Science Center) and Peter McGregor (Australian National University) used the Near-Infrared Integral Field Spectrometer (NIFS) spatial resolution over the spectral image cube.

Binary Stars and Bipolar Planetary Nebulae

The data reveal sharp density gradients, which are

Planetary nebulae frequently exhibit highly axisymmetric

interpreted to be regions where young massive stars

shapes as opposed to spherical symmetry. Binary star

have cleared some of their immediate surroundings

progenitors would naturally account for the angular

with radiation and winds. The kinematic signature of a

momentum these objects seem to require, but even

bipolar structure is evident in the Brγ emission (Figure

with large samples of planetary nebulae, few have

4), and this double-lobed structure is not aligned with

shown direct evidence for binary central stars.

with the Altair adaptive optics system to maintain high

The observed 11μm lensed images of the quasar H1413+117 (left), a model of the images, including two lensing galaxies (center), and the model residuals (right). The midinfrared data offer several advantages for studies of gravitational lenses, notably reducing the effects of microlensing by stars.

results from similar measurements at lower spatial resolution over larger physical scales. It is also not

Ph.D. student Brent Miszalski (Macquarie University and

centered on any of the detected point sources in the

Université Strasbourg) and an international team have

image or on the radio continuum sources. The bipolar

used GMOS on Gemini South, along with data from

flow could be associated with a lower-mass protostar

several other telescopes on the ground and in space, to

near the central ionizing source.

concentrate on post-common envelope nebulae. With this restriction on the evolutionary state of the nebulae,

Searching for Halo Substructure

almost 30% of the sample objects show direct evidence of bipolarity, such as the equatorial rings Figure 6 (next

Previous studies of the gravitationally lensed quasar

page) demonstrates. Moreover, after correcting for

H1413+117 identified flux ratios among the images that

projection effects, the authors identify up to 60% of

a single, central galaxy acting as the gravitational lens

the sample as bipolar. They conclude that the common

could not produce (although a smooth central galaxy

envelope phase is tied to the bipolar morphologies of



Figure 6.

The planetary nebulae M 2-19 and H 2-29 show prominent equatorial rings. The images are constructed from Ha + [NII] (red), [SII] (green), and [OIII] (blue} narrow-band GMOS images. Each image is 30 arcseconds across.

the nebulae. In addition, many of the sample members

to quantify the star formation history from past to

show signs of binary progenitors, exhibiting knots,


filaments, and jets. In addition, spectroscopy of these objects provides

Tracing Star Formation with Planetary Nebulae and HII Regions

metallicity indicators of each population when their stellar progenitors formed. The similar abundances of the HII regions and planetary nebulae indicate that little

Stars themselves are not the only tracers of star formation.

enrichment has occurred between the epochs of their

Laura Magrini (INAF - Osservatorio Astrofisico di

formation (Figure 7). However, dispersion in each of

Arcetri) and Denise Gonçalves (Universidade Federal

the groups suggests inhomogeneous abundances across

do Rio de Janeiro) used GMOS observations of HII

this dwarf galaxy, in which small-scale enrichment

regions and planetary nebulae in the starburst galaxy

with limited mixing may be significant.

IC 10 to reveal its star formation history. While this

Figure 7.

Abundance ratios of HII regions (empty circles) and planetary nebulae (filled circles) in the dwarf starburst galaxy IC 10, compared with relationships derived from observations of blue compact galaxies (blue lines). There is no significant difference between the younger and older star-forming regions, which the HII regions and planetary nebulae indicate. This also implies that little enrichment has occurred over the galaxy’s evolution.

galaxy’s current rate of star formation is high, older

Nancy Levenson is Deputy Director and Head of Science at

stellar populations are also evident. The current

Gemini Observatory and can be reached at:

work takes advantage of the different phases of stellar

evolution HII regions and planetary nebulae represent



by Christian Marois & Michael Liu

Exoplanet Imaging with the Gemini Telescopes Humanity has been dreaming about other worlds for more than 2,000 years. This dream partly became reality 400 years ago when Galileo Galilei used his 1.5-inch diameter telescope to discover the first “mini-solar system”: Jupiter and its four major moons. This discovery had profound philosophical implications, as it supported the Copernican model of the universe and showed that other worlds did not necessarily revolve around the Earth. Thus, the foundations of observational astronomy are intimately connected with the drive to find and study other worlds. This drive to understand our place in the universe continues to be one of the most compelling reasons to build new telescopes and instrumentation. The past 15 years have been an especially rich time, starting from 1995 with the discovery of the first planets around Sun-like stars through precision radial velocity measurements. While radial velocities have been a boon for identifying planetary systems, measurements of them are fundamentally indirect, since the planets are detected only indirectly via their gravitational influence. Direct imaging of exoplanets is the next frontier, as detection of their photons opens the door to a wealth of physical information not otherwise available.

The Contrast Challenge The Jupiter system is relatively easy to see with a handheld 1.5-inch telescope, due to its proximity to Earth and the modest brightness difference between Jupiter and its major moons. However, seeing planets around other stars is a far larger challenge. Planets are thousands to billions of times fainter than stars, because planets have small radii and emit very little light on their own. On top of this fundamental problem, there are several practical complexities of observing exoplanets from Earth. To begin, stars are located millions of times farther away than Jupiter, resulting in a very small angular separation between the stars and their planets—a representative scale is about one arcsecond to a few arcseconds. Also, the turbulence of Earth’s atmosphere blurs light as it travels from outer space to ground-based telescopes, causing the images of stars to be much larger and thus blend even more with the light from planetary companions. Finally, light is further blurred and scattered once it arrives into telescopes and their associated instruments, such that the central star’s dazzling bright light easily overwhelms any light from associated planets.



Altogether, the combination of this small angular

much cooler temperatures than stars, and thus they

separation and enormous brightness difference between

emit most of their light in the infrared.

a star and a planet can be described as both an angular resolution and contrast problem. Large telescopes and

Opening a New Era

complex instruments are required to separate the light of the star and planets in order to resolve the system,

The hunt for exoplanets around nearby stars by direct

and clever techniques are needed to suppress the

imaging started in the 1980s. But it really grew into

residual bright starlight and image faint planets in orbit

a serious effort in the 1990s, as large 4- to 10-meter

around other stars.

ground-based telescopes were equipped with firstgeneration AO systems and infrared cameras. These

From Galileo Galilei to Now







exoplanet detection around a Sun-like star found After 400 years of hard effort and innovation,

by radial velocity measurements (51 Peg b in 1995).

astronomers are now overcoming these challenges and

Indeed these radial velocity discoveries provided new

achieving the capability to see other worlds. Since

momentum to search for planets by direct imaging,

Galileo’s time, ever-larger telescopes have been built,

now that we knew gas giant planets definitely existed

placed on better observing sites, and equipped with

and were common around other stars.

more powerful instruments. In the 20th century, two major developments stand out as key to direct imaging

The first attempts at direct imaging with AO confirmed

discoveries of exoplanets. In the 1950s, Horace Babcock

that it would still not be an easy task. Even with large

suggested a revolutionary solution to correct the

telescopes and dedicated instruments, imperfections in

blurring caused by Earth’s atmosphere: rapidly change

telescope and instrument optics as well as the residual

the shape of a small mirror in order to cancel the

effects of atmospheric turbulence not corrected by AO,

atmospheric fluctuations in real-time. This technology

still inhibited the ability to probe the small angular

is now known as adaptive optics (AO).

separations around stars where planets were thought to be located. As a result of these challenges, much

The first generation of AO systems was installed on

effort was spent on conceiving and implementing new

astronomical telescopes in the early 1990s, and AO is

techniques to improve the contrast. As a result, three

now nearly ubiquitous on new large telescopes. In fact,

primary strategies have been deployed to boost the

the first science instrument at Gemini Observatory was

power of existing telescopes: coronagraphy, spectral

Hokupa‘a-36, an AO imaging camera developed by the

difference imaging (SDI), and angular differential

University of Hawai‘i and deployed on Gemini North;

imaging (ADI).

it was used for several programs that searched nearby stars for brown dwarf companions, the high-mass

Coronagraphy: the coronagraph, invented by Bernard

analogs to exoplanets. This system was subsequently

Lyot in 1939 to look at the Sun’s corona, is a relatively

replaced in 2003 by Altair, a more permanent higher-

simple and ingenious optical setup that can be adapted

order system built by Canada’s National Research

for high-contrast imaging of exoplanets. A small mask

Council’s Herzberg Institute of Astrophysics (NRC-

is placed at the instrument focal plane to physically

HIA), which is now routinely used in conjunction

block the on-axis light from the star. Most of the star’s

with other instruments on Gemini North on many

remaining light is then scattered to wide angles and

observing nights.

further blocked by a complementary mask downstream. The net result is that the central star’s light is greatly

The other key innovation was the development, in

suppressed, making it easier to detect faint objects in

the 1980s, of infrared (IR) array detectors, which

close proximity. Due to its ease of implementation, the

replaced the single-channel photometers used since

Lyot-style coronagraph is now a widespread feature of

the 1960s. Similar to the detectors found in today’s

imaging instruments for current AO systems.

digital cameras, IR detectors are sensitive to light at


infrared wavelengths and deliver digital images that

Spectral Differential Imaging (SDI): SDI is an

can be processed by a computer. The ability to detect

observing technique based on the very different surface

IR photons is very important, since exoplanets have

temperatures between a cool planet and a hot star. Since


planets are not sufficiently massive to sustain nuclear

sequence of observations. All the noise induced by

fusion, planets steadily cool with time from their

the telescope and instrument defects then stays fixed

initial hot state at formation. After about 100 million

with time relative to the instrument detector, while the

years, gas-giant planets are cool enough that many

image of planet revolves in the field of view around the

complex molecules can form in their atmospheres—a

bright star. The images are then analyzed to subtract

process that does not occur for ordinary stars. These

the static noise and retain the planet’s flux. ADI was

molecules include the formation of gaseous methane

first successfully used for science in the early part of

(CH4), which produces very strong features. One of

this decade by the two of us independently, M. Liu at

the strongest is at 1.6 microns, where the planet flux is

the W. M. Keck Observatory and C. Marois at Gemini

bright at wavelengths bluer than 1.6 microns and then

North (where ADI imaging was actually one of the

sharply drops to nearly zero at redder wavelengths. In

first science observations made with the Altair AO

contrast, stars have nearly blackbody spectra, smoothly

system). Recent improvements in the post-processing



reduction methods made by David Lafrenière have

simultaneous imaging at two (or more) wavelengths

further increased the power of this technique (the

adjacent to this sharp 1.6 micron drop to counteract the

“LOCI” algorithm). ADI was the key method used

time-variable images produced by the Earth’s turbulent

for the discovery of the HR 8799 multi-planet system,


first detected in 2007 with the Gemini North telescope





(Figure 1). Subtraction of the two spectral imaging channels removes the bright (methane-free) glare of the target stars and reveals any faint, ultracool (methane-bearing) planetary companions. Since the images at different wavelengths are acquired simultaneously, the images of the bright star suffer the same atmospheric AO residual and instrument aberrations at both wavelengths, producing much better subtraction than if the images were acquired non-simultaneously. The TRIDENT near-infrared camera was the first instrument to test

Up to now, three large-scale direct imaging surveys

this approach, installed at the Canada-France Hawai‘i

for exoplanets have been completed, or are ongoing, at

Telescope on Mauna Kea in 2001 as part of the Ph.D.

Gemini. The first one, the Gemini Deep Planet Survey

thesis work of one of us (C. Marois, under the direction

or GDPS (led by David Lafrenière, then at the University

of René Doyon, Daniel Nadeau, and René Racine at the

of Montreal), was carried out between 2004 and 2007.

University of Montreal). It has also been employed for

A total of 86 of mainly GKM stars have been acquired

the SDI camera at the Very Large Telescope and is one

with ADI for an hour with Altair/NIRI at Gemini

of the key features of Gemini’s new NICI instrument,

North. No planet was found, but the survey placed

described in more detail below.

very good constraints on the presence of Jupiter-like planets in wide (30-300 astronomical unit (AU)) orbits.

Angular Differential Imaging (ADI): ADI is an

GDPS concluded that less than 9% of GKM stars have a

observing strategy that uses Earth’s rotation to separate

3-12 Jupiter-mass planet between 50-250 AU.

the light of a planet from the light of a bright star. It thereby allows very accurate star light subtraction

An extension of the GDPS survey, called the

by post-processing in software. Large 8- to 10-meter

International Deep Planet Survey (IDPS), led by

telescopes are usually built such that the telescope

one of us (C. Marois) began in late 2007. The survey

rotation axes differ from the Earth’s axis, leading to

combines the use of the Gemini Telescopes, Keck, and

rotation of images as the telescope tracks an object in

VLT to acquire 1-2 hour ADI sequences of 100 nearby

the sky over time. To compensate for this field rotation,

young, high-mass (A- and F-type) stars. The main

telescopes usually employ an instrument rotator. With

goal is to remove the late-type bias of the GDPS and

ADI, the rotator is turned off, causing the instrument

study planet formation for a wide range of stellar host

to be perfectly aligned with the telescope for an entire

masses. This survey made a breakthrough discovery


Figure 1.

Before (left) and after (right) ADI processing of Altair/ NIRI HR 8799 data acquired in September 2008 at K-band. The left panel shows a 30-second exposure filtered to remove the smooth radially symmetric halo, while the image to the right is the combination of 60 ADI-processed 30-second exposures. The three 7- to 10-Jupiter mass planets are easily visible in the ADIprocessed images, while they were completely hidden by scattered light from the primary in the raw image. The solid white line in the right panel represents a one arcsecond angular separation. The concentric circles are the orbits of Jupiter, Saturn, Uranus, Neptune, and Pluto, viewed face-on at the distance of HR 8799 (39 parsecs or 127 light years). The central part of the images are masked out due to detector saturation. Both images are displayed with the same linear intensity range and have a 5.6 x 5.6 arcseconds field of view. North is up and east is left.


for high-contrast imaging, minimizing the wavefront distortions from the atmosphere, telescope, and instrument. The heart of NICI is its dedicated, advanced AO system, based on an 85-element curvature-deformable mirror developed by the University of Hawaii’s AO group. In addition, NICI employs all three of the observing methods described above—coronagraphy, SDI, and ADI—making it the first such instrument to use all these powerful

Figure 2.

Speckle subtraction using NICI’s dualchannel cameras. The left and center panels are images of the same star taken at the same time at slightly different wavelengths. The difference between the two images is shown in the right-hand panel. The darker circle in the center is due to the coronagraph (which allows about 1% of the starlight through so the precise location of the star can be determined). This example is only one of many exposures that were taken of this star, with the cassegrain rotator stationary, so any possible real planets would appear to rotate around the star in subsequent exposures. For more details on this technique (called ADI) see page 45 for descriptions of the ADI technique.

in 2008 when the HR 8799 planetary system was first

techniques to perform a large exoplanet imaging

imaged at Gemini (see Figure 1, previous page). The

campaign. As a result, NICI achieves excellent contrast

survey is still ongoing and follow-up observations are

and sensitivity to planetary companions around a large

being done on other candidate planets. (Each candidate

number of nearby stars.

needs to be acquired at two epochs to confirm that the object is in orbit around the star and not an unrelated

Looking Ahead

background object.) Over its history, Gemini Observatory has played a leading

The NICI Planet-Finding Campaign

role in the hunt for planets by direct imaging: deploying state-of-the-art AO systems, helping astronomers to

Gemini’s latest exoplanet finding campaign is the NICI

develop innovative high-contrast observing strategies,

Planet-Finding Campaign, being carried out at Gemini

and pursuing major dedicated surveys to search a large

South and is led by one of the authors (M. Liu).

number of targets. This continuous support, and the

The survey started in December 2008 and is in the

use of powerful instruments and algorithms, led to the

process of imaging about 300 nearby young stars. The

breakthrough discovery of the HR 8799 multi-planet

NICI Planet-Finding Campaign assembles a large-scale

system in 2007. And the major surveys being carried

coherent science program with a unified set of goals,

out now at both Gemini North and South telescopes

observing methods, and data analysis techniques. This

promise to further stretch our horizons about planetary

is the first such large official campaign being carried

systems around other stars.

out by the Gemini Observatory, and, accordingly, the NICI program is also intended to serve as a model for

In addition to the two vigorous AO surveys currently

even larger Gemini observing projects carried out with

ongoing, Gemini is also building a next-generation,

future instruments.

fully optimized, exoplanet-finding instrument called the Gemini Planet Imager (PI Bruce Macintosh). This

The campaign is based on Gemini’s latest facility

Aspen-derived instrument is scheduled for deployment

instrument, the Near-Infrared Coronagraphic Imager

in 2011. It is designed to detect and fully characterize

(NICI). This is a powerful new AO instrument and, in

Jovian planets located between 5 and 100 AU around

fact, is the first instrument for any large (8- to 10-meter)

nearby young stars. Gemini is definitely at the forefront

telescope designed from the outset for direct detection

of this exciting emerging field!

of extrasolar planets through high contrast imaging. NICI was built by Mauna Kea Infrared (MKIR) in

Christian Marois is a research associate at the National Research

Hilo, Hawai‘i, with Doug Toomey as the principal

Council Canada Herzberg Institute of Astrophysics. He can be

investigator (PI), Christ Ftaclas (IfA/Hawai‘i) as the

reached at:

project scientist, and Mark Chun (IfA/Hawai‘i) as the lead for the AO system.

Michael Liu is an astronomer at the Institute for Astronomy at the University of Hawai‘i. He can be reached at:

NICI was designed as a complete end-to-end system



by Dennis Crabtree

Time-Exchange Programs Gemini, which prov ides telescope access for seven national communities, has a user base with very broad science interests. However, instruments for 8- to 10-meter telescopes are expensive. This makes it very challenging for Gemini to provide instrument capabilities that meet the full scientific needs of our diverse community. Other 8- to 10-meter telescopes also face this challenge to various degrees, and in addition may not have access to both hemispheres. For example, Keck decommissioned their mid-infrared instrument several years ago. Also, the W. M. Keck and Subaru observatories do not provide access to the southern skies for their communities. One obvious solution to this problem is to establish time exchange agreements between various facilities so that users can get access to a broader range of instrumentation. Gemini has been very active in providing access to an expanded instrumentation suite for our community by establishing time exchange programs with Keck and Subaru. (See the companion article by Tom Geballe on the early Gemini time exchange program with Keck (NIRSPEC) following this article.)

Time Exchange with Keck Gemini’s time exchange program with Keck provides access to the HIRES—the high-resolution echelle spectrometer ( This program began in semester 2005A. HIRES is a very powerful and successful instrument that provides spectral coverage between 0.3 and 1.0 micron with resolutions between 25,000 and 85,000. In exchange for access to HIRES, the Keck community can apply to use MICHELLE or the Near-Infrared Imager and Spectrometer (NIRI) on Gemini North or the Thermal-Region Camera Spectrograph (T-ReCS) on Gemini South. While time exchanges are simple in concept, putting them into practice is a bit more complicated. Since Keck runs exclusively in classical mode, our agreement with Keck is to simply exchange classical nights. The base number of nights exchanged is five nights per semester, but can be lower if one partner’s demand is lower.



Then there is also the issue of sky brightness to agree

Kyoto (see page 52), more than one person noted that

upon. Since Gemini is giving the Keck community

they were discouraged from applying for the Subaru

access to infrared instruments, this is essentially bright

time because of the small number of nights available.

time. However, HIRES is an optical instrument and

Second, the instruments available on Subaru have

some of the science requires, or at least would benefit

been increased to include most of Subaru’s instrument

from, being scheduled in dark time. Keck is very


protective of their dark time, so the Gemini HIRES

Spectrograph (HDS) (

nights are scheduled in either bright or gray time, with

Instruments/HDS/hds_umev111.pdf ), a high-resolution

the exact windows in the semester being advertised as

echelle spectrograph providing a resolution of 95,000

part of the Call for Proposals.

with a 0.4 arcsecond slit. Wavelength coverage is from






0.3 – 1.0 micron. In recent semesters, the actual number of nights exchanged with Keck has been less than five nights as

Gemini is working to formalize the time exchange

a result of low demand for Gemini time from within

agreement with Subaru, and we hope to increase the

the Keck community.

amount of time exchanged to 25 nights. The amount of time exchanged will necessarily be limited by the

Time Exchange with Subaru

demand from each side, with either partner able to establish a lower number of nights for a given

Gemini’s time exchange program with Subaru has

semester if their community doesn’t express sufficient

evolved since it began in semester 2006B. Initially,

interest. We hope to have an agreement in place with

Gemini and Subaru exchanged 50 hours of queue time

Subaru before the 2010B Call for Proposals.

(service mode on Subaru) with the Gemini community having access to SuprimeCam (a wide-field optical

Lessons Learned

imager) and MOIRCS (a near-infrared wide-field imager for which multi-object spectroscopy (MOS)

Time exchanges can be a very effective approach for

capability became available in later semesters).

providing instrumental capabilities to the Gemini scientific community that are not available at Gemini.

Subaru is classically scheduled, with specific nights

As in any barter agreement, each side must have

set aside for service-mode observing. This is not the

something that the other side considers of interest. In

same as the Gemini queue, as the observing conditions

this context, Gemini can provide:

cannot be guaranteed, and the observing time may be completely weathered out. Gemini observers, used to

• access to both hemispheres;

our queue model, expected that they would receive

• excellent mid-infrared capabilities at both sites;

Subaru data taken in the conditions specified in the

• laser guide star adaptive optics at Gemini North;

proposal, and problems arose when this didn’t happen.

• multi-conjugate adaptive optics at Gemini South

After two (three for SuprimeCam) semesters of trying


to exchange Gemini queue time for Subaru service

• cryogenic infrared multi-object spectrograph at

time, both sides realized that this was impossible with

Gemini South (future).

the two very different operations models. Beginning with semester 2008A, the time exchange program with

While we are working to expand the time exchange

Subaru reverted to the Keck model of exchanging

with Subaru, we are always exploring possible time

classical nights for classical nights with five nights

exchanges with other 8-meter telescopes. These

being exchanged.

strategic time exchanges need to be kept in mind when developing future instrumentation plans. It may

For semester 2010A the time exchange with Subaru

be more effective to exchange time to get access to an

has been expanded significantly. First, there are

instrumental capability than it is to build a similar

5-10 nights available depending upon demand. The

instrument for Gemini.

possibility of having access to 10 nights of Subaru


time may increase the number of proposals from the

Dennis Crabtree is Gemini’s Associate Director of Science

Gemini community. At the Gemini Users’ Meeting in

Operations. He can be reached at:


by Tom Geballe

Gemini Community Access to Keck in 2000-2001: Early Time-Sharing In 1999, the Gemini and W. M. Keck observatories agreed to a unique exchange of assets. Gemini provided Keck with a science-grade Aladdin-I 1024 × 1024 InSb array for Keck’s medium- and high-resolution near-infrared spectrograph, NIRSPEC. In return, Keck allocated three nights of observing with NIRSPEC per semester to the Gemini community for the next two years, beginning in semester 2000A. Science fellow Marianne Takamiya and I were given the task of conceiving, developing, and carrying out, on behalf of the Gemini partnership, an observing program that would utilize those twelve nights. At that time, Gemini North was just entering its operations with a paucity of instrumentation, and Gemini South was still under construction. Gemini data were only slowly seeping out to the Gemini community. Marianne and I proposed to model the Gemini/NIRSPEC Program after the highly successful United Kingdom Infrared Telescope (UKIRT) Service Observing Program, in which investigators could easily apply for small amounts of observing time within a few weeks of the observing dates. We hoped that such an approach would give many astronomers across the community their first experience with infrared data from a large telescope. Our suggestion was accepted by Gemini Science Committee and recommended to then Gemini Director (Matt Mountain), who gave us the go-ahead. We familiarized ourselves with NIRSPEC, created a website (still accessible at: and an application form, assembled a small group of external scientific assessors, prepared to technically assess the proposals, advertised the program to the community, and awaited the response. On the website we provide the following description of the program: “The primary selection criteria for the proposals are scientific quality, technical feasibility, efficiency, and observing time requirements... The primary



Figure 1.

refereed papers come from four of the seven partner

The K-band NIRSPEC spectrum of the Lynx gravitational arc (z = 3.357), reported by Fosbury et al. (2003, ApJ, 596, 797). Analysis shows that this remarkable arc is an H II galaxy containing roughly one million O stars, magnified by a factor of ~10 by a complex foreground cluster of galaxies.

countries: the U. S., UK, Brazil, and Australia, from Gemini staff, and also from Korea. The list of refereed publications can be found on the Gemini website at: keck.html These papers report observations of a wide range of objects: z = 6 quasars, gravitationally lensed galaxies at high redshift, globular clusters, old metal-poor stars, young stars, and solar system objects. Two examples are shown in the accompanying figures. Figure 1, from intent …is to obtain science for the international Gemini

a paper by British astronomer Robert Fosbury (of the

community, which has many members and interests. It

European Southern Observatory) and collaborators

is expected that most proposals will request from one-

shows the infrared spectrum of a galaxy at a distance

half hour to a few hours of integration time. Unless

of 11 billion light years that is undergoing an intense

a long program is judged to be superior to several

episode of star formation. As viewed at that distance

shorter programs, it is unlikely to be done first.”

such an event normally would be unspectacular. However, in this case the star-forming galaxy lies

The response from the community was overwhelming.

behind a cluster of foreground galaxies, whose gravity

A total of 179 proposals were received during the two-

amplifies the light coming toward us by an order of

year period, covering a wide range of topics. The


roughly 45 proposals per semester were far more than could be accommodated in three nights. Typically the

In contrast, the data in Figure 2, measurements of the

15-20 highest-rated programs were tentatively accepted,

ratio of the two stable isotopes of carbon, are for stars

and only about half of these were attempted. The choice

in our own backyard, that is within the Milky Way.

of which of these to observe was based on scientific

The rarer isotope, 13C, is manufactured deep in a star

priority, weather conditions, and brevity.

and is mixed to the surface of the star during the star’s red giant phase, reducing the observed value of 12C/13C.

Figure 2.

The mixing is believed to proceed in two stages and

A plot of 12C/13C versus absolute stellar magnitude for red giants in four globular clusters, showing different values of 12 13 C/ C above and below the bump in the luminosity function (BLF). Gemini/NIRSPEC data are denoted by triangles. These observations, reported by Shetrone (2003, ApJ, 585, L45) are the first time that the predicted decline of 12C/13C, due to extra mixing at the BLF, was detected in globular clusters.

these data, reported by University of Texas astronomer Matthew Shetrone, are the first time that the predicted decline of 12C/13C due to the second stage of mixing, at the so-called “bump in the luminosity function” in red giants, was detected in globular clusters. The Gemini/NIRSPEC contributions to this study were the observations of NGC 6528 (denoted by triangles) and clearly were key to establishing the result, because, unlike the data from two of the other three globular clusters, the NIRSPEC data included red giants both above and below the bump. Now, eight years after the end of the program, we


can perhaps at last tally the final results. Despite poor

One two-hour NIRSPEC observation of the 3-micron

weather on four of the 12 nights, and several hours

spectra of Saturn and its largest moon Titan in

of good weather lost to technical problems, high-

November 2001 has resulted in four papers by an

quality data were obtained for roughly 25 programs.

international team headed by Professor Sang Joon

At present, at least 14 refereed papers (the most recent

Kim of Kyunghee University in Korea. Three of these

of which appeared in early 2009) and at least as many

concerned Titan. The first reported the discovery

contributed papers and posters have resulted from the

of mesospheric line emission by several species of

Gemini/NIRSPEC Program. The first authors of the

hydrocarbon. The second refined and expanded the


analysis of the Titanian spectrum and provided much additional information on chemical abundances in the upper and lower atmosphere, as well as on the altitudes of cloud decks and haze layers. The third (published in 2009) reported the detection of additional molecular bands, for which laboratory data had been unavailable earlier. In






although limited to one instrument, can be considered a forerunner of queue observing and Directorâ&#x20AC;&#x2122;s Discretionary Time at Gemini. On each night of the program, sky conditions as well as scientific priorities influenced the observersâ&#x20AC;&#x2122; on-the-spot decisions as to which program to observe. Advances in astronomy clearly require time-consuming observing projects as well as quick looks. Nevertheless, it is interesting to note how programs such as this one, involving only short projects, can result in high publication rates compared to normal classical or queue observing. The success of the Gemini/NIRSPEC program was due in large part to the support of the staff of the W. M. Keck Observatory. The program is also indebted to Professor Ian McLean of UCLA, who built NIRSPEC and provided helpful advice to the observers, and to the referees who volunteered many hours of their time: Beatriz Barbuy, Antonio Chrysostomou, Malcolm Coe, Fred Hamann, Tim Heckman, Robert Howell, George Mitchell, Keith Noll, Steve Rawlings, Kris Sellgren, Verne Smith, David Tholen, Chris Tinney, and Sylvain Veilleux. Tom Geballe is a staff astronomer at Gemini North. He can be reached at:



by Jean-René Roy, Atsuko Nitta & Nancy A. Levenson

Joint Gemini/Subaru Science Meeting & 2009 Gemini Users’ Meeting Set among the historical setting of the old capital city of Japan, Kyoto, about 200 users and staff of the Gemini and Subaru observatories and of the National Gemini Offices participated in the first joint Gemini and Subaru science conference, from May 18 to 22, 2009. The meeting was held in the Clock Tower Centennial Hall of the University of

Figure 1.

Participants at the 2009 Gemini/ Subaru Science Meeting in Kyoto, Japan.

Kyoto. The five-day meeting provided a unique opportunity for participants to share a wide range of research topics spanning studies of our solar system and exoplanets, super-metal poor stars, cosmology and the high redshift universe. Users also discussed innovative observing techniques and unique collaborations, while they looked toward the future of both observatories and at how we can work together to best serve very demanding users while remaining competitive. In particular, methods and procedures for coordinating the use, planning, and construction of future instruments were discussed, as well as expanding the exchange of observing time, which would take advantage of the complementarity of the strengths of Gemini and Subaru (see articles on time exchange programs starting on pg. 47). The conference concluded with the 3rd Gemini Users’ Meeting on Friday, May 22, when Gemini staff updated users on issues related to doing science on Gemini and provided a forum for user input. Several Japanese



Figure 2.

astronomers attended the Gemini Users Meeting as

Noboru Ebizuka of Nagoya University shares a poster with Gemini’s Bernadette Rodgers.

well. Members of both communities expressed a strong desire to see the observing exchange program between Subaru and Gemini expanded significantly to a larger number of nights per semester with all instruments of both facilities made available, which has begun for 2010A. (Through 2009B, the exchange was for five nights/semester, using two instruments; now up to 10 nights may be exchanged, with Gemini user access to six Subaru instruments.) dramatic

Buddhist temples in the beautiful northern hills of

announcement that Gemini’s financial participation

Kyoto. The evening banquet on Wednesday, May 20,








was a lively get-together, and many participants had

(WFMOS) would not proceed as planned, following

the opportunity to get introduced to lesser-known

the cancellation of the WFMOS Project by the

aspects of Japanese culture. Before and after the

Gemini Board. Details on this announcement can

meeting, several attendees took vacations to discover

be found in the Gemini Board WFMOS Resolution

more about this fascinating country.




at: Gemini’s Director, Doug Simons, made the announcement

The scientific and logistical success of the joint

at the opening of the conference and emphasized

conference was in large-part due to the efforts of the

the fact that future collaborations between Gemini

scientific and local organizing committees (SOC and

and Subaru are still a core foundation for the two

LOC). The SOC was led by two co-chairs Masashi

observatories. “While I’m extremely disappointed

Chiba (Tohoku University) and Timothy C. Beers

about the status of Gemini’s participation in

(Michigan State University). The LOC was chaired

WFMOS, we are still committed to a partnership

by Kouji Ohta (Kyoto University). On behalf of all

with Subaru to best serve our user communities,”

the participants and users of the Gemini and Subaru

said Simons.

Telescopes, we also wish to acknowledge the support provided by the members of the SOC and LOC: Masashi Chiba (Tohoku University) Co-Chair Toru Yamada (Tohoku University) Motohide Tamura (NAOJ)

Figure 3.

Gemini’s Jean-Rene Roy (now at the NSF) speaks at the meeting’s banquet.

Kazuhiro Shimasaku (University of Tokyo) Yoshiko Okamoto (Ibaraki University) Kouji Ohta (Kyoto University) Timothy C. Beers (Michigan State University) Co-Chair Isobel Hook (Oxford University) The two main goals of the conference were to promote

Chris Packham (University of Florida)

a mutual understanding of both communities and

Scott Croom (Sydney University)

to highlight the international nature of modern

Marcin Sawicki (St Mary’s University)


Local Organizing Committee (LOC)

These goals were clearly achieved through the

Kouji Ohta (Kyoto University) CHAIR

presentation and discussion of new ideas, the

Hajime Sugai (Kyoto University)

fostering of collaborations, and the building of

Tomonori Totani (Kyoto University)

new friendships. The sharing and interactions

Hideko Nomura (Kyoto University)

were helped by several activities such as a tour of

Atsuko Nitta (Gemini Observatory)



Figure 4.

The two previous Gemini Science Conferences were

Participants at the Gemini/Subaru Science Meeting banquet.

held in May 2005 in Vancouver, Canada, and June 2007 in Foz do Iguaçu, Brazil. Jean-René Roy is Senior Scientist at Gemini Observatory and is currently on leave of service in the Large Facility Project Group at the National Science Foundation. He can be reached at: Atsuko Nitta is a Gemini science fellow at Gemini North. She Finally, we wish to highlight the masterful leadership

can be reached at:

and careful management of Kouji Ohta and his team at the University of Kyoto for a very smooth

Nancy Levenson is the Gemini Deputy Director and Head of

meeting: in particular the cautious behind-the-scene

Science at Gemini South. She can be reached at:

handling of the threatening uncertainty that the H1N1

pandemic was casting on the conference. Luckily the University of Kyoto did not shut down (as was possible), and we used their excellent conference facilities to the very end of the conference and for the 2009 Gemini Users’ Meeting.



by Dennis Crabtree

On-site Observing with the Gemini Telescopes When most observers hear of on-site observing, they automatically think of visiting the telescope as a classical observer. While Gemini supports classical observing runs, 90% of the scheduled time is in queue mode, which does not require anyone from the program team to be present. However, having a program in the queue does not preclude a visit to the telescope and participation in the execution of your program. A visiting observer for a classical run will typically come a day or two before their observing run starts and leave almost immediately after the end of the run. They may be at Gemini long enough to give a colloquium, interact with a few staff members, and perhaps go out for lunch. Of course, nothing stops the visiting classical observers from staying longer if they wish! Gemini welcomes visits from either the principal investigator (PI) or co-investigators (co-I) on queue programs. Ideally, the visit will be scheduled during the period of the year when the queue program is most likely to be observed. We encourage queue visitors to come to Gemini for a period of at least a week, perhaps even longer if they are students. A typical visit by a queue observer will include the visitor talking with various Gemini staff to learn how the Gemini queue is populated, optimizing their Phase II with their contact scientist, and learning a bit more about how Gemini really works. It is very worthwhile for a visitor to spend time with the queue coordinator on duty to understand the factors that determine how the nightly plans are populated. Gemini will make every effort to ensure that the visitorâ&#x20AC;&#x2122;s program is in the queue during the period of the visit. Of course, the program will not be executed unless the observing conditions match those required by the program. The visitor will then spend a few nights at the summit to see night-time operations. Hopefully, he or she will get to see their program being executed (and most likely run the telescope for these observations) and have



a chance to look at the data to ensure everything is

who visit Gemini South also get a chance to sample

working as planned. One of the benefits of visiting

local Chilean culture and activities.

during queue operations is that the visitor gets an opportunity to see other instruments being used

Once visitors have data for their programs, they can

for a variety of science programs. And the night

get advice on data reduction or plans for follow-up

may be livened up by a “slew now” GRB Target of

observations with the same instrument or perhaps

Opportunity observation!

another of Gemini’s instrumentation suite. Students are encouraged to give a science lunch talk or a

In our experience, the students associated with

colloquium during their visits.

queue programs who visit Gemini are the ones who benefit the most. The students get a chance


to experience the complete range of workings of a

students, have found their visits to be a rewarding

modern observatory. They can spend time talking







to their contact scientist about their program, talk with others on staff with similar science interests,

For details on arranging visits to either Gemini site

participate in colloquia and science coffees, learn

please see:

about our instrument development (particularly







they visit Gemini South) and participate in queue observations for their programs and others. Those

Dennis Crabtree is the Associate Director of Science Operations and located at Gemini South. He can be reached at:

The visit to Gemini South was a unique and valuable experience. Being a beginner in near-infrared spectroscopy and the only Phoenix user at University of Toronto Astronomy, it was difficult for me to follow all the instrument details and data acquisition requirements at the beginning. During my visit to Gemini South, I worked closely with the instrument scientist German Gimeno to learn about details of Phoenix data reduction and analysis. I also worked on the summit for six nights and participated in queue observations with various instruments, including Phoenix, GMOS, and T-ReCS. During the Phoenix observations, I learned about the data acquisition process, which was extremely helpful for me to better understand the intricacies of the observations. I also learned about the instrument capabilities of GMOS and T-ReCS which may be useful in future research. Furthermore, I interacted with many Gemini staff and visitors, gaining alternate perspectives on my project. The visit allowed for an opportunity to absorb much more information than would be possible during a short visit associated with a classical observing run. Sherry Yee Ph.D. student – University of Toronto



by Eric Tollestrup

Planning for Gemini’s Future Instruments

As Gemini approaches its 10th anniversary, we are reminded that many of the instruments that shared first light with the telescopes—the first-generation instruments—have reached obsolescence and need to be replaced. Moreover, as we all know from our everyday experiences, technology evolves at an ever-increasing pace. This is also true for astronomy-related technology too, and these new technological advances beckon us to enrich our scientific capabilities through newer, better instrumentation. Therefore, at Gemini we have

Figure 1.

Gemini North’s instrument cluster, showing the current generation of instruments (without NIFS).

embarked on a new round of selecting the next generation of instruments (the fourth) to replace our older, more cantankerous, though inanimate friends. So, how is this instrumentation selection done? Who is involved? The short answer to the latter is, “all of us in the Gemini community.” How the selection process happens is an interesting story that everyone in the community should know about so they can participate in the decision-making. After all, without great instruments, a telescope is not much more than a wonderful expression of modern engineering prowess.



For an insightful perspective, a review of the past is

was later refined into a set of recommendations

instructive. In May 1994, the Gemini Board approved

by the first International Gemini Instrumentation

the first generation, or Phase I, instruments:

Workshop, held in Abingdon, England, in January 1997. The “Abingdon” instruments that became

Mauna Kea:

reality were the Laser Guide Star mode for AO, the integral field unit (IFU) for GNIRS, NICI, NIFS,

• Optical Acquisition Camera;

FLAMINGOS-2, and eventually Multi-Conjugate

• 1- to 5-micron Imager;

Adaptive Optics (MCAO)/ Gemini South Adaptive

• 1- to 5-micron Spectrograph;

Optics Imager (GSAOI).

• Multi-object Optical Spectrograph; • 8-to 30-micron Imager (shared);

In November 2003, the Gemini Board approved

• Canada-France-Hawai‘i Telescope (CFHT) Fiber

key aspects of the Gemini Aspen Instrumentation


Workshop, which led to Gemini’s third-generation, or “Aspen,” instruments. This generation of instruments

Cerro Pachón:







challenging suite of observing technologies seen • Optical Acquisition Camera;

thus far. By 2011 we should see the fruits of this

• Multi-object Optical Spectrometer;

endeavor with the commissioning of the Gemini

• High Resolution Optical Spectrograph (HROS);

Planet Imager (GPI).

• Shared Instrument with CTIO; • 8- to 30-micron Imager (shared).

Along the way there were additional visitor instruments,





The resulting instruments included the Near-

FLAMINGOS-1, CIRPASS, and TEXES, all of which

infrared Imager and Spectrometer (NIRI), Gemini

helped fill in missing or niche capabilities. The

Near-Infrared Spectrometer and Imager (GNIRS),

bottom line is that Gemini and its community have

Gemini Multi-Object Spectrometer (GMOS (N and

always striven to provide the best instruments in an

S)), High-Resolution Optical Spectrometer (HROS

effort to capitalize on the outstanding capabilities of

(later to be bHROS)), Thermal-Region Camera

our two 8-meter telescopes.

Spectrometer (T-ReCS), acquisition and


cameras (not intended for science but GMOS-N/S

As you may have noted, most of the instruments

fulfills the need for optical science cameras) and,

were initiated in (and therefore based on technology

due to limited budgets, the shared mid-infrared

from) the 1990s. As we approach the second decade

instrument MICHELLE (from the United Kingdom

of the 21st century, it is time to retire or upgrade

Infrared Telescope (UKIRT)) and 1- to 5-micron

these senior citizens from the past millennium. The

spectrometer Phoenix

first step is already underway. The Gemini Science

(from the National Optical for

Committee (GSC), which reports to the director and

bHROS, all of these instruments are still actively

advises the Gemini Board, has been developing a

used, though bearing the burdens of old age. These

straw man list of possible instruments or capabilities

first instruments were envisioned as “workhorses.”

as part of the long range planning for Gemini. This

Adaptive optics systems were also part of the facility

list is not definitive or exclusive, but is intended

development, from which Altair resulted.

as a starting point to initiate discussions at the





grassroots level—i.e., they are developing “talking In the June 1996 issue of the Gemini Newsletter,

points.” Though the GSC has informally polled

then-director Matt Mountain first presented to the

members of the community, and in this way has

Gemini community the initial thinking on the Phase

established a reasonable list of possible instruments.

II instruments, where the emphasis was on upgrades

It is up to the community to build a consensus for

to the Phase I instruments and on instruments

the final recommendations for future capabilities or

that “targeted unique capabilities that the Gemini

instruments for Gemini.

telescopes offered for instrumentation.” This thinking



Next, over the course of 2010, each of the partner

The timing of this process is intimately connected

countries, lead by each National Gemini Office

to the next five-year budget and Gemini Cooperative

(NGO), will sponsor Town Hall-type meetings at

Agreement (2011 to 2015) with the National Science

their respective national astronomy conferences.

Foundation (NSF), the executive agency that oversees

These are intended to foster grassroots discussion

Gemini Observatory. By having a proposed set of

from the community, and help determine each

next-generation instruments ready for approval

partner country’s priorities and needs. The first of

by the November 2010 Gemini Board meeting,

these Town Hall meetings for the U. S. community

the observatory can implement these instruments

will be held as a special session at the American

within the context of the next 2011-2015 budget and

Astronomical Society winter meeting held in early

associated Cooperative Agreement.

2010 in Washington, D. C. Similar meetings are intended for the Royal Astronomical Society National

So like an electioneer who cries, “If you care about

Astronomy Meeting in April, the CASCA meeting in

the future of your country, vote!”, we implore to

May, the Astronomical Society of Australia Meeting in

you, as a Gemini user, “If you care about the future

July, and the Brazilian Astronomical Society Meeting

of Gemini, participate in Gemini Instrumentation,

in September. These meetings are your opportunity

the Next Generation!”

to let your NGO know what is important for you as an observer on Gemini’s telescopes. It is important to

Eric Tollestrup is Associate Director of Development at Gemini,

emphasize that Gemini Observatory representatives

located in Hilo. He can be reached at:

will be attending each of these Town Hall meetings to answer your questions and better understand the various subtle nuances that might arise. Since Gemini will have to successfully execute any final recommendation, it is very important to us to truly understand our community’s desires. Other meetings or avenues for input are possible, and interested Gemini community members should contact their NGOs to offer suggested input venues that would help this grassroots process. Already, some communities are conducting web-based surveys to gather information. One-on-one discussions with your NGO are encouraged too. As with a wedding, this is your chance to “speak now or forever hold your peace!” The culmination of these grassroots efforts will be the third International Gemini Instrumentation Workshop (aka, the “A” meeting in reference to the previously mentioned Abingdon and Aspen workshops), during the middle of 2010. The intent of this third workshop is to merge together the interests of all the partner countries into a single set of recommendations to the GSC, which in turn can fine tune this into a recommendation to the director and Gemini Board for approval at the November 2010 board meeting.



by François Rigaut & Céline D’orgeville

MCAO Progress

Two significant milestones were achieved over the past couple of months on the Gemini Multi-Conjugate Adaptive Optics System (GeMS) laser infrastructure.

Figure 1.

The laser service enclosure on the Gemini South telescope (silver box on altitude platform).

The laser service enclosure (LSE) is the laser clean room designed and fabricated by Gemini to house the Gemini South laser for the GeMS project. The LSE was installed on the Gemini South telescope on August 13, 2009, changing the appearance of the facility from that of its twin in the north (see Figure 1). The large (8.5 × 2.5 × 2.5 meters) structure was mounted onto the new support structure that had been added to the side of the telescope about a month earlier. The LSE and its support structure add another five tons to the existing 340-ton weight of the telescope. A careful survey of the telescope, and laboratory testing of the LSE, ensured that the laser clean room could be installed in a single day without disrupting night-time operations. Performance of the 8-meter Gemini South telescope is unaffected. Another significant milestone was reached on September 4, 2009, when Lockheed Martin Coherent Technologies (LMCT) delivered their sodium laser system to the W. M. Keck Observatory headquarters in Waimea. The Keck laser is the first of two such laser systems to be procured by Association of Universities for Research in Astronomy (AURA) on behalf of both the Keck and Gemini Observatories. Factory acceptance testing of the Keck laser in Colorado took place in July 2009, followed by about a month of retooling to address a few shortcomings. These were mainly software-related. Performance testing of the Keck laser demonstrated unexpectedly high output power at nearly twice the required 20 watts for this system, and near diffraction-



limited beam qualities with M2 in the 1.1- to 1.2-micron

On July 27, in a ceremony at the offices of the

range. Short-term and long-term laser performance

Dirección General de Aeronáutica Civil (DGAC,

stability, which is of critical importance to GeMS

the Chilean equivalent of the US Federal Aviation

operations, is within specifications. This augurs well


for future Gemini South laser performance. Factory

negotiation, AURA signed the umbrella agreement

acceptance of the 50-watt Gemini South laser system

that covers the collaboration between AURA and the

is expected to occur in Colorado before the end of

DGAC in support of the safe operation of laser guide

the year, and delivery is scheduled in early 2010.

star systems at the AURA observatories in Chile.






This agreement achieves an important milestone in Things are also progressing well on the CANOPUS

our progress toward operating the variety of laser



systems that are coming to Cerro Pachón in the

re-implementation of the static tomography high-

near future, including (in probable order of initial

level software, we were able to solve the tomographic

use) Andes LIDAR, the Gemini Multi-Conjugate

compensation of the static non-common path

Adaptive Optics (MCAO) system, and the SOAR

aberrations, and reach Strehl ratios of 96% +/- 1% at

ground layer adaptive optics system.






1.65 microns simultaneously over the entire output field of view. The specification was 94% +/- 3, so we

In parallel, we are progressing in our negotiations

are clearly better. This shows that we understand the

with VIA56, a company DGAC has been working

system, its calibration and non-linearities fairly well.

with to implement similar applications. Using data

It also proves that there is very little aberration (50

from Chilean radar coverage, the software provided

nanometers or less) outside of the altitude range (0

by VIA56 gives the location of any aircraft in our

to 9 kilometers) correctable by the three deformable

area. A side benefit is that it allows us to have some


advance notice of a possible laser shutter, which gives us some amount of flexibility to finish an exposure

As announced in the last GeminiFocus GeMS update,

and/or to switch to an alternate program before we

we went through a maintenance and upgrade of

have to shutter. Our All-Sky Camera (ASCam, see

the CANOPUS real-time computer last June,

earlier GeminiFocus issues) still provides a backup

working with an engineer from the Optical Science


Company (tOSC). We were able to fix a number of bugs and implement long-needed features. The

Finally, we have had some help for the last few

contract with tOSC is now essentially closed. The

months from German Fernandez, an electronics

real-time computer is performing well, and has been


extensively tested while it has been used almost

Argentina, who has been working half time (two

daily for the past year and a half.

weeks per month) on the electronics for the new






cooling system on the MCAO bench. Concerning high-level operational software, our high-level engineering interface, MYST (MCAO

François Rigaut is the Adaptive Optics Senior Scientist and

Yorick Smart Tools) has also reached stable status. In

located at Gemini South. He can be reached at:

the coming months, we will implement operation

features in MYST, such as high-level loop control and sequencing, high-level status checks, and adaptive

Céline D’orgeville is the Gemini South Senior Laser Engineer.

optics performance optimizations. The Observing

She can be reached at:

Tool modifications to accommodate GeMS and Gemini South Adaptive Optics Imager (GSAOI) are also essentially complete (at least for a basic set of functionalities, more features will come with the mid 2010 release), and feature user-friendly multi guide star selectors, and other features.



by Sarah Blanchard

The Greening of Gemini Question: How much energy does Gemini consume? Simple Answer: A lot. First, consider electricity. In an average month, Gemini North consumes 232,600 kilowatt-hours (kWh) and Gemini South uses 246,000 kWh of electricity. Based on an average cost of U. S. $0.56 per kWh at Gemini North and an average cost of 82.29 Chilean pesos (about U. S. $0.14) per kWh at Gemini South, that translates to a total annual cost of nearly two million dollars and a contribution of more than 3,500 metric tons of CO2 to the Earth’s atmosphere. Chile enjoys relatively low electricity costs, but energy costs in Hawai‘i are among the highest in the United

Figure 1.

States. When energy prices soared in February 2008, the cost of a kWh in Hawai‘i hit $1.07, pushing Gemini

Cyclists like astronomer Rachel Mason appreciate the bike racks at both Gemini North and Gemini South Base Facilities.

North’s electricity bill above $240,000 for that month alone. Then there’s travel. During 2008, Gemini staff completed 616 trips by air, totaling nearly five million miles— the equivalent of almost 11 round trips to the Moon. That amount of air travel produced a carbon footprint of 2,426 metric tons of CO2 in 2008. (A carbon footprint is the amount of greenhouse gases produced as the direct and indirect results of a human activity, usually expressed in equivalent tons of carbon dioxide. Generating electricity, using jet fuel, producing products or food—the carbon footprints for all these activities can be measured and tracked. See for more information.) Given those sobering statistics and our strong desire to be good stewards of our planet, we had to ask: How can Gemini reduce energy usage and manage our organization in a more sustainable way?



Planning for Sustainability

Setting the Stage for Success

In mid-2008, faced with soaring energy prices and

Adopting “green” standards and behaviors requires

a growing awareness of our need to reduce energy

commitment, a willingness to look critically at the

use, Gemini Director Doug Simons asked Facilities

way our facilities are presently operating, and an

Manager Peter McEvoy to begin an observatory-wide

eagerness to find opportunities for improvement.

energy-reduction project by launching the “Gemini

Gemini’s answer to these challenges has been an

Green Blog,” an internal forum that asked Gemini

enthusiastic “Yes!”

employees to submit their concerns, suggestions, and solutions for reducing our energy consumption







and conserving resources.

Committee (EPOCC) first convened in January 2009 to lead both projects. Members of the EPOCC

The staff responded enthusiastically. Green Blog

include observatory director Doug Simons and

suggestions ranged from the large and complex

several employees from the departments with direct

(“install photovoltaic solar roof panels”) to the

responsibility for managing energy purchase and

simple and obvious (“set printer defaults to print on

consumption: engineering, facilities, information

both sides of the paper,” and “remind everyone to

systems, administration, and finance.

turn off lights when they leave their offices”). The committee’s initial goals were to: The Green Blog suggestions were grouped into categories: technology investments, transportation


reflect Gemini’s commitment to energy efficiency;





develop an energy plan and mission statement to implement obvious and immediate energy-

saving actions; •

create strategies for continuous improvement;

make changes that are acknowledged to be

Figure 2.

Facilities Specialist David Moe sorts the recyclables at the Hilo Base Facility.

the “right thing to do,” even if they don’t result in immediate cost savings to Gemini (e.g., recycling, proper disposal of hazardous materials, sourcing products produced from renewable energy sources, requiring vendors to follow green practices).

Immediate Improvements After we analyzed the recommendations in the materials reduction. These staff recommendations

Green Blog, we found several areas for immediate

became the foundation for developing the plan for

and simple improvements. Over the past six months,

an observatory-wide 2009 Energy Initiatives project,

here’s what we’ve accomplished.

aimed at reducing energy consumption through specific actions.

Reducing off-hours energy use in both base facilities. Although Gemini runs 24/7/365, not all of

A second director’s-level Band One project—Energy

our office equipment needs to do so. We conducted

Planning Oversight and Control—was also launched

energy audits to review what’s plugged in and when,

to address the ongoing requirements for long-term

and what pieces of equipment can be unplugged or

energy monitoring and conservation. Together,

turned off when not in use. We found more than

these two initiatives reflect the need for immediate

400 pieces of electrical equipment plugged in at each

solutions as well as overall changes to the way

site, and more than a hundred turned on although


they were not being used. These initial findings


conserves energy.




suggest we might save about U. S. $10,000 a year by



turning off all unneeded electrical office equipment.


Simply identifying and communicating this issue to

system to monitor resource needs and power

our staff has helped us reduce the number of items

down unneeded servers without negative impacts

left plugged in when they’re not required.

to applications or users. With virtualization, work






demands are balanced and consolidated among We’ve tested and installed motion sensors to

networked servers and work stations to provide

automatically turn off lights in some areas at the

greater efficiency, saving kWhs and CO2 in required

South Base Facility (SBF) and on Cerro Pachón, and

cooling. Virtualization also provides significant

we’re in the process of installing motion sensors

additional savings

in other areas, both at Gemini North and Gemini

maintenance costs. Our progress in this direction


will include further user consultation and careful

in physical hardware and

implementation. Our Information Services Group (ISG) is also helping

system that provides system-wide information such

Improving the efficiency of the air conditioning and heating systems at the base facilities. We installed programmable

as uptime and usage statistics. These data will

thermostats in office areas at the Hawai‘i base

help us track redundant machine use, highlighting

facility (HBF) and programmed the air conditioning

where further off-hours savings can be made. (This

units to run only as needed. (Estimated savings: 200

system also saves time and effort, making it easier

kWh per day.) To help our Hawai’i staff members

for ISG to deploy computer configurations, software

stay comfortable during hours when the AC is off,

installations and patch management.)

HBF has also purchased several portable fans for

us create energy-saving efficiencies by introducing new technologies and a new computer-monitoring

office use. At Gemini South, the same practice has produced similar savings, and we’ve also purchased several portable heaters for anyone working outside normal office hours during the winter months in La Serena. Two older air conditioning units at HBF have been replaced with newer, more efficient models. (And we’ve received more than a thousand dollars in rebates from Hawai‘i Electric Light Company as an added bonus.) Bids are being developed for similar chiller upgrades at SBF.

Maximizing printer and computer efficiency. By installing transparent printer services for UNIX

Improving the  efficiency of the air conditioning systems at the telescopes.

and Windows users, ISG has reduced the need to

One of the biggest challenges has been to identify

install multiple printer drivers, and duplex printers


can be automatically configured to print two-sided

telescopes, which consume 70 to 80 percent of

as the default option. This service also helps monitor

Gemini’s total electricity needs. Air currents caused

printer usage. Gemini South Network System

by warm air inside the dome and cooler air outside

Administrator Chris Morrison notes that in 2008,

can play havoc with local seeing conditions, and can

68,412 pages were produced by the printers connected

damage instruments through thermal changes and

to the universal printer server. (Approximately eight

condensation. So the temperature of the telescope

trees were required to create those 68,000 pages of

itself, the equipment and instruments, and the


inside of each observatory must match the expected







night-time temperature. Humans and equipment


Another major energy-saving step is to consolidate

working inside the observatory create heat, so air

Gemini’s network resources through virtualization,

conditioning is vital. And, anything with a large







We’ve also standardized disposal procedures for

must also be temperature controlled. Therefore,

electronic and hazardous waste, such as computer

observatories demand a great deal of energy for air

equipment and non-rechargeable batteries. We’ve


purchased rechargeable batteries and set up batterycharging stations at HBF, SBF and the two telescope

At Gemini South, we’ve implemented a new

facilities to provide AAA, AA, C, and D batteries for

program for the device that controls the way the

portable tools, flashlights, and wireless equipment,

observing floor is cooled at the beginning of the

such as computer mice and keyboards.

observing night. This is expected to produce a double benefit of significant annual energy savings

Reducing energy used in transportation.

as well as contributing to improved local seeing

Gemini South has limited its number of vehicle trips

conditions at the beginning of the night. Its impact

between the base and the mountain by developing

is being studied for consideration at the Gemini

a collective transport solution that involves the

North telescope too.

use of a 37-seat passenger bus (one round-trip per day, mainly for day crew), and a combination of

Ironically, the harsh winter of 2008-2009 at Gemini

a 9-seater Mercedes passenger van and a double-

North caused our energy use (and costs) to drop

cab Ford F-150 (for use with light cargo and in bad

dramatically when compared to the

weather or heavy terrain conditions),

previous winter, because the snow

now encouraging staff to limit travel

As Gemini director used by all staff. Doug Simons has stated, Gemini North is working with Mauna Kea Support Services and “Energy is simply too other Mauna Kea observatories to the feasibility of a similar precious a resource to determine shuttle service that would provide use without a careful, transportation from base facilities to the Hale Pokahu lodge. Additional deliberate approach to solutions are also being considered to transportation costs and the managing it.” Besides, reduce impact of vehicles on the mountain, this is simply the right while still maintaining high standards of safety for our staff. thing to do.

whenever possible, increase their

Bicycle racks are in regular use

use of interactive technologies for

outside the offices at both Gemini

and ice prevented the dome from being opened on many nights—thus reducing





needed to provide air conditioning.

Increasing our employees’ awareness of the energy costs in travel. Effective astronomy research



collaboration, and travel between our two telescope sites by key personnel is especially important to Gemini’s continued success. However, we’re

collaborative communication, and try to reduce

North and Gemini South base facilities as well (see

Gemini’s overall travel carbon footprint. We maintain

page 62).

a “Gemini carbon footprint database” which is updated every month to inform travelers about the

LEEDing the Future

carbon impact for each trip. These recent initiatives represent definite process Promoting efforts to reduce, re-use, and recycle.

improvements, but we recognize them as piecemeal



at this stage—just a good start, according to our

practical at both summit and base facilities for paper,

initial plan. We know we have a long way to go

cardboard, returnable beverage containers, batteries,

on a steep learning curve, as we integrate this

plastic, and metal. We encourage staff members to

forward-looking philosophy into the organization.

bring cups, mugs, plates and utensils for use in the

So, we are reviewing and fine-tuning the advances

office, to eliminate, whenever possible, the use of

already made, as we consider opportunities for

disposable kitchenware.

additional improvements. Ultimately, we want to







see “green” practices become simply “business as


usual,” regardless of whether energy prices rise, fall,

Buildings as a roadmap for our energy conservation






or remain the same.

and sustainability efforts in the future, using their very detailed guide to help us plan Gemini’s further

So, to help us formalize our long-term energy


strategy, Gemini has become a corporate member of the U. S. Green Building Council, the organization

Sarah Blanchard is the Administration and Facilities Team

that administers the Leadership in Energy and

Leader at Gemini North. She can be contacted at:






Rating scheme. We plan to use the LEED program

THE GEMINI GREEN MISSION STATEMENT: “Exploring the universe, caring for Earth” “Explorando el Universo, cuidando la Tierra” GEMINI’S GREEN PRINCIPLES: •

Gemini aims to be one of the most energy-efficient major astronomical facilities, with the highest standards of environmental

protection. •

As part of Gemini’s purpose to “explore the universe and share its wonders,” we are responsible and accountable for our

actions. We will carefully consider and protect Earth’s valuable environmental resources, while fulfilling our commitment to science and applying solid financial principles to our operations. •

Gemini is committed to operating responsibly in the interests of our communities, local and global, now and always. Our

corporate citizenship promotes sustainability through effective, efficient operations and positive environmental stewardship. GEMINI WILL: •

ENGAGE our employees’ inventiveness and enthusiasm for a green working environment.

INCLUDE energy efficiency and environmental considerations as an integral part of our business planning and decision-

making processes. •

MEASURE our energy efficiency and environmental achievements.

COMPLY with all relevant legislation and, wherever appropriate, seek to adopt more stringent standards of energy efficiency

and environmental excellence. •

REDUCE consumption whenever possible, as we constantly assess our use of goods and services, travel and communications.

USE recycled products wherever practical.

RECYCLE, whenever possible, acquired goods and consumables at the end of their useful life.

PROMOTE the highest possible standards of energy efficiency and environmental protection among our suppliers and

partners. •

COMMUNICATE regularly with our employees and stakeholders to increase everyone’s understanding and support for

energy efficiency and environmental protection.



by Neil Barker

Broadening Participation at Gemini G e m i n i  h a s  a l w a y s  t a k e n seriously its important role in the global astronomical community, as well as the need, through education and outreach, to inspire current and future generations in the wonders of the universe. Figure 1.

Thanks to our international partnership, our two remote locations in Hawai‘i and Chile, and our hiring practices, today’s diverse Gemini staff originates from 20 countries around the world. Against a background of changing demographics, however, our future success depends upon inspiring the next generation from the broadest and most diverse audience possible. This will allow us to continue achieving excellence in our organization and the scientific community as a whole.

Vincent Fesquet works with intern Lucia Poláková as part of the Akamai internship program in Hawai‘i.

During 2009, Gemini embarked on an initiative sponsored by the Association of Universities for Research in Astronomy (AURA) to broaden participation in our future workforce. Historically, women and individuals from minority groups, and those with disabilities, have been underrepresented in astronomy. Gemini has been very successful at attracting female astronomers who make up about 30% of our Ph.D. science staff. Furthermore, the Gemini senior leadership team originates from three continents and has 50% female representation.



As Gemini’s Recruiter, I have enthusiastically accepted this role on behalf of Gemini. The Diversity Advocate’s role is envisioned as taking an overview of the human resource and outreach elements that affect under-represented minorities, women, and persons with disabilities. 2. The Gemini staff has also participated in an AURA Employee Climate Survey. This survey was generated across all AURA centers to improve our understanding of the employee climate. The survey has provided valuable feedback on perceptions of the climate within each institution. The results of the survey have been communicated to Gemini staff and action plans have been developed.

Figure 2.

Gemini South Science and Engineering staff planning for nighttime operations.

The objective of this initiative is to strenghten

In April of this year, a Broadening Participation



in Astronomy Workshop in Tucson, Arizona,

communities within our future workforce, as well as




brought together management, human resources,

the smaller, or geographically-dispersed institutions,

and outreach representatives from all of the AURA

that use our facilities. This is not a straightforward

centers. The workshop featured presentations from

exercise, with diversity issues being woven deeply

leading scientific organizations, including NASA and

into the political, social and economic fabric of

the NSF. Each presentation provided insight into

society. However, this initiative, which Gemini has

meeting the challenges of developing a future with

embraced, is integral to the success and future of

a diverse workforce. The workshop culminated in

astronomy, especially in a world that is competing

a presentation to AURA member representatives

for increasingly scarce resources.

which shared thinking from the workshop and some initial goals and asked for input from the

How are We Going to Achieve This?

member representatives present. Participants left the workshop inspired with many great ideas and the

1. AURA has published an extensive action plan

feeling that there was a real opportunity for change.

to highlight the principles, goals, and strategies

There appeared to be overwhelming support for a

for broadening participation at AURA facilities

call to action.

see: A workforce and diversity committee, made up

Following this inspiring and motivating conference,

of experts in diversity and education, has been

detailed planning has begun in order to develop,

established to advise and support AURA. Within

understand, and crystallize our goals and plans

Gemini and the other AURA centers, a diversity

related to diversity over the next five years.

advocate has been appointed to be the focal point for coordination of local Gemini and AURA initiatives.

We have been asking some important questions in this process:

Figure 3.

Gemini team at the Tucson Broadening Participation conference.

How do we address the broadening participation

initiative leveraging Gemini’s unique international partnership and the location of our twin telescopes within Hawai‘i and Chile? •

What will the future international workforce look




What will attract people to work in astronomy

and specifically to work at Gemini? •

How do we work with educators and professional

bodies to inspire more people to pursue careers in astronomy and related fields? •

How do we better target under-represented

groups within the United States and more broadly through our international reach? •

What role can Gemini and other centers play

in this through outreach and work experience opportunities? •







expanded? •

How can we provide an environment that will

balance employee and observatory needs in the future? These are just some of the questions we are considering as we embark on this journey. A closing remark from one of the member representatives at the Tucson conference resonates in my mind: “We can all talk about the need to change, but only through actually changing what we do will we achieve our goals.” For more information please contact Neil Barker (address






Participation Website at: diversity/ Neil Barker is the Recruiter at Gemini North. He can be reached at:



by Peter Michaud

Looking Ahead With Your Help

In March of 1992, the Gemini Project Newsletter was born with the rather matter-of-fact opening sentence: “This is the first newsletter from the Gemini Project.” As we approach the 20 year of publication, it th

is time to re-evaluate the future of Gemini’s key communication tool to our various communities. As the montage above reveals, the Gemini Newsletter has evolved dramatically during its lifetime. From a very utilitarian collection of photocopied black-and-white pages, to the full-color publication you are looking at now, the Gemini Newsletter has matured in lockstep with the observatory. Today, GeminiFocus, the name adopted for the newsletter in 2005, is a publication produced by our staff and supported by our users; it is the observatory’s flagship publication.



Of course, every publication has to plan for the

The staff working on GeminiFocus appreciate your

future, and the future of any print publication

input and time to help us plot this future. We hope

looks very dynamic. As Gemini explores ways to

this publication will continue to serve your needs as

communicate more effectively, efficiently, and with

a user, associate, or interested layperson.

a greener global impact, we need you, our readerâ&#x20AC;&#x2122;s input to determine the best way forward. To that end, we have created a web-based survey to collect opinions on the evolution of this publication (see: If the web survey isnâ&#x20AC;&#x2122;t able to capture your opinion, please send a note to the GeminiFocus Managing Editor, Peter Michaud at:



by María Antonieta García

Gemini’s Rapa Nui Connection

The connection between Gemini Observatory and Easter Island (or Rapa Nui) began last year with the creation of an astronomy booklet (Cuadernillo Astronómico, Volume 1) that included information on the profound relationship between the Chilean Indians and the cosmos. A few months later, when the International Year of Astronomy began, Gemini Observatory—

Figure 1.

Students, teachers, tourist guides, hotel staff, and local firefighters prepare to enter the StarLab® portable planetarium after participating in a FamilyAstro workshop.

together with the University of Concepción and the Anthropological Museum of Father Sebastián Englert— coordinated the first astronomical cultural exchange with Easter Island. The events, occurring from August 1 to 9, included staff lectures, presentations in the Gemini StarLab® portable planetarium, solar observing, and nighttime stargazing. In addition, the FamilyAstro program provided engaging activity/inquiry-based learning opportunities for island families and teachers.



Figure 2.

“It’s good that the students listen to what is actually

Students from 3rd to 12th grade from four Rapa Nui schools receive educational materials after enjoying a StarLab® planetarium program.

going to happen [in the sky] and thus value what we have the chance to witness,” said Kava Calderón Tuki, a local elementary school teacher, who attended a presentation in Gemini’s StarLab planetarium. Other events included a guided tour of the millenary lava caverns (hosted by staff of the Anthropological Museum), a chance to witness the sunrise at the Ahu Tonga Riki, and a lecture given by local archeaologists on the most recent discoveries made in Rapa Nui. More than 700 people, about a third of the Rapa Nui population, participated in the program’s activities. In the same week, an important soccer match was played on the island, and there was also an ardent political debate in the background, which without a doubt also appealed to the interests of the islanders. Having so many people show up for our festivities


annexed in 1888. (It is currently governed

as a province of Chile’s Valparaiso Region.) The island culture is of Polynesian origin, and residents call their home Rapa Nui, or Te Pito Te Henua, which means “The World’s Belly Button.” The island was originally colonized around the 4th or 5th century AD by the Ariki (King) Hotu Matu’a and a group of followers, who came from the island of Hiva, in

was an accomplishment given the competition.

the Marquesas Islands. The island was discovered

The director of the museum, Francisco Torres,

reported seeing it on Easter Sunday in 1722. That

referred to the Polynesian origins of the island as he

by Europeans when the Dutchman Jacob Roggeveen discovery date led to the island’s current (common) name.

Figure 3.

María Antonieta García is the Outreach and Media Specialist at Gemini South. She can be reached at:

Rapa Nui students study the sun with a Sunspotter® solar viewer while a University of Concepción student explains total solar eclipses, like the one that will be visible in July 2010 over the island.

discussed the importance of the astronomy event. “The first sailors who arrived down here were guided by the stars as an only tool,” he pointed out, “which is a very important detail that also unites us with Hawai‘i.” Nearly 200 kilograms of educational materials were transported for the event as a donation by Lan Cargo. This allowed us to offer the people of Easter Island an opportunity to learn more about astronomy before the total eclipse of the Sun they will be able to see in July of 2010. Easter Island is located 3,700 kilometers from continental Chile, and is a special territory of



by Peter Michaud

An Outreach Conjunction

Figure 1.

Participants in the annual Kona Teacher Workshop, a Journey Through the Universe westside event for science educators.

A conjunction of a unique astronomical sort occurred on October 3, 2009, at the Natural Energy Lab in Kona, Hawai‘i. This wasn’t a conjunction of astronomical bodies per se, but an alignment that was a long time in the making, and it shone brightly for those who participated. This conjunction brought together several dozen very motivated Hawai‘i Island teachers, an impressive stack of GalileoScopes (an International Year of Astronomy product, see:, a team of astronomy educators, and sponsors that included other Mauna Kea observatories and local businesses. The day-long workshop was organized by Gemini Observatory as an extension of the Journey Through the Universe program that traditionally focuses on the east side of the Big Island. Janice Harvey, coordinator of the event, provides some background on the program, “For some very complex reasons, the Journey Through the Universe program ( is limited to East Hawai‘i (Hilo area),” she said. “So West Hawai‘i (Kona region) gets the short end of the stick. We do this workshop to help address this imbalance.”








Figure 2.


Nancy Tashima (Onizuka Space Center) and Tom Chun (Kamehameha Schools) building their GalileoScopes.


provided comments as part of an evaluation, and excerpts throughout this article capture the enthusiastic reaction of the participants. Based on the evaluations, it appears that the east/ west scales were well on their way to achieving a better balance. The 2009 workshop is the second annual Kona teacher event organized by Gemini and, based on its success, will not be the last. “This year was special because we were able to introduce the GalileoScope to the teachers, and they loved it. We really got to teach some wonderful optics and astronomy,” said Tim Slater of the University of Wyoming, Wyoming Excellence in Higher Education, Endowed Chair in

“I really appreciate Gemini’s outreach to teachers. We’re fortunate to live near the planet’s prime astronomical location, and I thank you for sharing knowledge and materials with the students of the Big Island. Our students are a great investment.” Joanna Yin, Na‘alehu Elementary

Science Education, who led the teacher activities. According to Janice Harvey, “This was the first GalileoScope teacher workshop to include a 128-page teacher’s resource and activity book produced by Tim and his group at the University of Wyoming.” Each teacher received a copy of the book and a completed GalileoScope

“Lately, professional development has been limited and could be costly. I appreciated the expertise of the presenters, the hospitality, including lunch, and the materials.”

Figure 3.

Rhanda Vickery from Waikaloa Elementary prepares to build her GalileoScope.

Kristin Tarnas, Hawai‘i Preparatory Academy

(sponsored by Gemini) as part of the workshop. Gemini wishes to thank the other Mauna




participated (especially the Thirty

“And I saw the craters on the moon from my house!!!! Thanks!!!!!!!” Debbie Low, Kealakehe Elementary

Meter Telescope), and Big Island Toyota for providing lunch and refreshments. The Gemini North outreach staff has a similar workshop in East Hawai‘i planned as part of the “flagship” Journey Through the Universe program in February 2010. Peter Michaud is the Public Information and Outreach

“Thank you for organizing such a fantastic workshop and for giving us that wonderful Galileo Telescope…I can hardly wait to introduce my students to it.” Phyllis Cabral, Connections Public Charter School

Manager at Gemini Observatory. He can be reached at:



by Peter Michaud

Australian Student’s Image Revealed “I thought they were pulling my leg,” said Australian high school student Daniel Tran when he first found out that his entry into the Australian Gemini Partner Office imaging contest was selected as the winner. His target, known as the Glowing Eye Nebula (NGC 6751), entered the Gemini South queue for 2009A, where it received one hour of telescope time for multi-band imaging with the Gemini Multi-Object Spectrograph. The image is shown here as a color composite produced by Travis Rector from Daniel’s data. The contest, sponsored by the Australian Gemini Office solicited high school students from across Australia to submit a target and explain why it would make a good image. According to Daniel, when he saw the object online at the WorldWide Telescope

Figure 1.

(WWT, a project of Microsoft), he found that “its unique color

Image of NGC 6751 obtained for the Australian Student Imaging Program. Inset image shows the presentation on September 23, 2009. Holding the photo are (left to right), PAL College (high school equivalent) teacher David Lee; Christopher Onken from the Australian Gemini Office; and PAL College student Daniel Tran, who was responsible for the winning entry. Photo: David Marshall.

and structure made me want to know more about it, and the name itself caught my attention and started to reel me in.” Note: see the Gemini image gallery on the WWT at: The image was revealed at an event on September 23, 2009, hosted by Christopher Onken (Mount Stromlo Observatory) and Dr. David Frew (Macquarie University), who explained to the students the context of the image and how this planetary nebula glows due to its central star at the end of its life. The image will appear on the cover of the January 2010 issue of the Australian version of Sky and Telescope. The Australian Gemini Office (AusGO) also announced that the second iteration of this program is now being planned for 2010.



by Joy Pollard

Bringing her contagious smile and warm laughter to the Gemini offices, Nancy Levenson is a welcome addition to the Gemini familia (or ‘ohana in Hawaiian). As the new Deputy Director and Head of Science, Nancy assumed her role at Gemini South only a few months ago and

Introducing Nancy Levenson

already her influence is propelling Gemini forward. Nancy’s first introduction to Gemini established her as a mid-infrared user partnering with Gemini North astronomer Rachel Mason to probe the core of the active galaxy NGC 1068. Nancy’s research interests focus on active galactic nuclei (AGN). She is anxious for more opportunities to utilize Gemini’s mid-infrared capabilities,





wavelength observations, to help unravel the mysteries shrouded within AGNs. A Rhodes Scholar, Nancy graduated with bachelor’s degrees from both Harvard and Oxford Universities, completing both her Master’s, and Ph.D. in astronomy at the University of California at Berkeley in 1997. She conducted her postdoctoral research at Johns Hopkins University. Gemini users with access to new and improved In 2002, she joined the faculty of the University of

instrumentation. She also wants to better understand

Kentucky in the department of Physics and Astronomy,

the perspectives of partner countries and their needs in

where she gained first-hand experience with Gemini’s

selecting and funding new instruments.

optical/infrared capabilities and queue-based observing model. Nancy’s experience with Gemini equips her

When asked about future scientific research at Gemini,

with a strong user perspective and sensitivity to the

she says she prefers to defer to the many “brilliant”

needs of our science communities.

minds that have already been making these decisions; supporting a vision where Gemini plays to its well-

In addition, she has served on the Gemini Science

established strengths in infrared observations, and


developing adaptive optics capabilities.






understand the user’s perspective from representatives of the Gemini partner countries. She also served on

Since her move to Chile, Nancy feels she is finding her

the ALTAIR committee (2008-2009) that surveyed and

stride, and, as is obvious at the staff birthday parties, is

assessed the user needs of large telescopes in the U. S.

enjoying flexing her Spanish language skills! To relax from her frequently demanding workload she plans to

Nancy says she is excited about addressing one of the key

spend some time adapting to the clay tennis courts in

findings of the ALTAIR committee, namely, providing




by María Antonieta García

Gemini’s Adaptive Optics Chef François Rigaut Not 10 minutes after lighting the fire under his well-used

of flavors from his homeland in France. Eventually,

wok and beginning to cook, the aromas from François

he shares the name of his favorite restaurant in Paris,

Rigaut’s kitchen begin to flow. He’s totally concentrated

which, until now has been a well-kept secret; it is the

on his work—adding tomato, zucchini, onions, and

Brasserie Bofinger, located in the Bastille district.

eggplant to sauté—all to create ratatouille which, in addition to being the name of a popular animated

It has been 17 years since François last lived in France,

movie, is a delicious vegetarian dish that François likes

and during this time he has been fortunate enough to

to prepare for his family.

travel millions of kilometers around the world. His unique work locations have allowed him many special

(Opposite page) François Rigaut enjoys the tranquility of cooking in his kitchen.

“I’m the chef of the house,” he asserts proudly. François,

experiences. “Being able to watch lava flowing in

when he’s not cooking at home, is the Adaptive Optics

Hawai‘i is something really incredible,” he recalls. Once

Senior Scientist at Gemini South, and he makes it clear

while scuba diving near Hilo, he recalls encountering

that he works without any recipe whatsoever (when

a single dolphin that suddenly expanded into a pod of

cooking!) His repertoire includes an array of more than

them. “That was a wonderful sensation which,” he says

20 favorites, with curry and au gratin dishes being his

regretfully, “I only photographed in my memory.”

specialty. “I enjoy good cuisine,” he points out while savoring a sip of red wine, which evokes memories



François also savors and appreciates the more mundane

Photo by M. Paredes



or subtle experiences in life, such as the time on Paranal

Ali Khan who François first heard during an observing

when he became totally absorbed in the tranquility

run several years ago. Ali Khan’s qawwali music is based

of his surroundings.

in repetitive chants that are supposed to eventually take

“For the first and only time, I

experienced total silence,” he says. ”The sensation was

listeners into a trance.

so incredible that I could hear my blood beating in my veins. I have been very fortunate to experience all these

François’ first experiences in what would become


a long affiliation with Chile began in 1987 as part of France’s compulsory military service. It also offered the

François’ training is in astronomy, but he feels that he has

option of civil service abroad. This allowed François

a wider scientific viewpoint these days. “I don’t consider

to spend 16 months working with the site testing

myself as an astronomer anymore; I am a physicist,”

group at Observatorio La Silla. “This led me to install

he says, noting that he does not write astronomical

testing equipment in Cerro Vizcachas,” adding that “I

research papers anymore and he has devoted himself

went to Paranal while it was still [in] a very natural

professionally to the development of the field and theory


of adaptive optics. It is in this area where François feels he can build more concrete, tangible things than with

After fulfilling his French military commitment,

pure astronomical research.

François returned to France to embark on his Ph.D. in Instrumentation and Astronomy at the Université Paris

A French community called Limours, just south of

VII under the supervision of Pierre Léna. There, he

Paris, is where François grew up. He is the only child

had the good fortune to work on the final development

of a medical laboratory administrative assistant and

stages of the first adaptive optics system for astronomy

an engineer. His dad (the engineer) worked in the

called “COME-ON”.

Atomic Energy Commission, and François believes he inherited his inclination for science from him. Youthful

With his Ph.D. and experience in the rapidly emerging

experiences at an international school, where he mingled

field of adaptive optics, François was quickly hired

with a diverse, multinational group of friends, made

by the Science Research National Center in France.

him comfortable with many different kinds of people.

He relocated to the Canada-France-Hawai‘i Telescope

“Since then, I have coexisted in multicultural and multi-

(CFHT) in Hawai‘i, where he led the development and

ethnic environments all of my life,” he points out.

commissioning of PUEO (named after the Hawaiian owl with exceptional eyesight). François describes it as

François spent many years of his adolescence learning to

the first user-friendly adaptive optics system, allowing

play classical guitar. “I practiced between eight and ten

astronomers to use it with the push of a button. “It

hours a week, but I never considered it as a profession,”

has tremendously aided the work of astronomers who,

he recalls, while confessing that these days he barely

before its development, had to rely on three or four

takes the guitar out of its case more than a couple times

persons to do the work, which is now automated,” he

a year.


Although he doesn’t make his own music these days,

Gemini Director Doug Simons (also at CFHT during

François does appreciate the classical music of Johann

the same period as François in the early ‘90s) recalls,

Sebastian Bach, particularly the pieces from the Well-

“[PUEO] was a radical concept at the time and sparked

tempered Clavier—a collection that Bach wrote to show

great debates within CFHT. Was this technology really

that the unification of the musical scale works. “In the

mature enough to catapult [AO] into the realm of

past, musical instruments were not tuned with the

“reliable” facility-class instruments?”

current keys,” he explains. In the end, Simons concludes that the project was a


Bach is not the only musician François favors. He has

great success and contributed profoundly to astronomy.

an eclectic appetite for many modern genres, including

“PUEO proved that it’s possible to package/control

pop and rock music of the ‘70s and ‘80s. At yet another

an adaptive optics system that works with just a few

extreme, he enjoys the Pakistani musician Nusrat Fateh

clicks on a computer screen that the general astronomy


community can use. This was a big confidence builder

proposing ground layer adaptive optics (GLAO), a new

for others around the globe pondering the merits

approach to wide-field AO.

and risks of building similar systems amid a myriad

“I feel a sense of responsibility to Gemini and the user

of skeptics.” Simons continues, “I feel lucky to have

community,” François says. “I love what I do. I am one

just witnessed the beginning of this revolution on the

of those few persons that when Friday comes, I say,

summit of Mauna Kea nearly 20 years ago.”

‘Darn, it’s already Friday!’”

François’s early work on PUEO would eventually

María Antonieta Garcia is the Outreach and Media Specialist

have a profound impact on Gemini as well. However,

at Gemini South. She can be reached at:

before coming to work at Gemini, he was hired by the European Southern Observatory (ESO) as the AO Group Head. There, he discovered some of his strengths and also determined some his life’s priorities. “I was not made for management,” he confesses. “I figured management could come later in life, and I still wanted to do some research.” After two years at ESO he decided it was time for a life change, and he gave up what was a lifetime position in France to accept his position at Gemini so he could help lead the observatory’s nascent adaptive optics program to maturity. That began eight years of living in Hilo and working at Gemini North. During this period, François was responsible for what he describes as his most challenging projects; the completion and commissioning of the Altair AO system and integrating it with the laser guide star system on Mauna Kea. It was also in Kona, Hawai‘i, on Bastille Day, 2002 that he married Gemini’s senior laser engineer Céline D’orgeville. Today, the family lives in Chile, with two children (daughter Anaïs, 3, and Lucas, 1). François also remains close to his two teenage daughters (Natacha, 17, and Camille, 15) from a previous marriage. While his older daughters live in France, they have both visited Hawai‘i and Chile, making his family almost as global as himself. François’ “professional family” also includes students he has mentored. “I am proud to have trained several students who have now succeeded well in the adaptive optics community,” he said. Meanwhile, his professional life is still extremely demanding, with the development of the Multi-Conjugate Adaptive Optics (MCAO) system for Gemini South, which he calls “our baby.” (See the MCAO update on page 60 of this issue.) Among François’ other professional accomplishments, he is particularly proud of having written the first paper



by Peter Michaud

In Love with the Stars Dolores Coulson “I love stars, that’s what it’s all about.” With these words,

As a teenager, Dolores recalls being profoundly impacted

Gemini North’s lead System Support Associate, Dolores

by early visits to the Palomar Mountain Observatory. “I

Coulson, only hints at her passion and enthusiasm for

would sit on the stairs and say to myself, ‘How can I

her life’s work.

get in there, this is what I want to do!’” After getting an immediate response to an inquiry (with pictures) from Palomar astronomer Jesse Greenstein, her fate was sealed. “I was beyond the Moon,” she said. “That was it; I had to do it! [What Jesse did] was wonderful. It was such a nice thing to do to encourage someone.” But, Dolores had a very different encounter a few years later that could have literally bumped her off this path

(Small Image) Dolores stands outside the Palomar Mountain Observatory as a youth. (Opposite) Dolores today outside of Gemini.

to the stars. It all started when she was working in her hometown of Encinitas, California, as a typesetter at a local newspaper and she (characteristically) tried to help out a friend. That friend, a reporter, was writing a story on a new bullring in Tijuana. He asked Dolores if she was willing to have her picture taken in the bullring, but that it



Photo by K. Puâ&#x20AC;&#x2DC;uohau-Pummill



would require some training first, since the photos

She shook off the trauma of the bull stampede and

would involve a close encounter with a bull. As it turns

entered the ring for the rest of her commitment. But

out, the encounter was a lot closer than Dolores ever

the adventure wasn’t over. Her bull (which she suspects


was switched, or grew longer horns) threw her to the ground. She recalls from her shaken memory, “All I

“For six weeks, every weekend I’d go down there to

remember is all of these people running around and

practice,” she recalls. At that point, she got tired of losing

trying to keep the bull away. I even remember seeing

her weekends to the bullring and said, “I’m not doing

a clown!”

this anymore. Let’s do it now. I’m not going to get any better!” On the day that she was to be photographed in

In her usual fashion, Dolores tells this story with humor,

the ring, Dolores had to select her bull from the corral.

humility, and a lot of humanity, (“… there was no way I

“I picked the nicest, kindest, sweetest little bull with

was gonna hurt that bull… .”) But most of all, her story

the shortest little horns and very kind eyes,” she said.

shows her determination to succeed. She finished the photo shoot, and since then, her determination has been reflected in her professional achievements. Dolores came to Gemini in 1997, at a very critical time both for herself and the observatory. She transitioned to Gemini from her work at the Joint Astronomy Centre, which supported her pursuit of a graduate degree from the University of Central Lancashire (in the United Kingdom). Her thesis, titled “A Detailed Infrared Study of the Star-forming Region NGC 2071 IRS,” allowed her to consummate her love affair with stars and the starbirth process; all while she envisioned a creative new structure for staffing nightly telescope operations. This became the foundation for Gemini’s System Support Associate (SSA) model. Dolores says that today she is most proud of how that vision has been implemented at Gemini and the fact that SSAs are able to advance and broadly participate in observatory operations. “No other observatory has that,” she said. “I’m proud of that.” Being the first person hired by Gemini to establish the observatory’s nightly operational functions, and the telescope’s user interface, Dolores is able to share

Dolores has a much closer encounter with a bull than she ever anticipated...

However, while walking down a narrow corridor to join

a history that few current staff members can fully

her “kind-eyed” bull in the ring, she heard a commotion

appreciate. Nowhere is this more obvious than when

from behind. There was shouting (in Spanish), and

she describes the first nights when the telescope was

when she looked back she saw bulls stampeding toward

pointed skyward.

her in what looked like an impromptu running of the bulls! According to Dolores, “I made myself as thin

Reflecting on the early commissioning of the Gemini

as possible. I sucked in my stomach, I think it went

North telescope system, Dolores says, “We used to

through the other side of the building! The last thing I

observe from the observing floor, and things used to get

remember seeing was this huge bull’s horn coming right

so cold. Those were the days of old, the days of cold!”

at me. I closed my eyes and that was it. I don’t even


remember hearing anything, then I opened my eyes and

She recalls the uncoated mirror reflecting the Moon’s

they were all gone—and I was still alive; I had already

light and using a hand-panel to point the telescope. “It

said goodbye!”

was cold,” she said, “until we built a ‘homeless shelter’


on the observing floor. It was made out of plastic

Her close personal friend and colleague Colin Aspin

and everyone tried to get near the one heater in the

recalls how passionate she is about the Dallas Cowboys


football team, partly because of the big star on their helmets! Colin tells of a Superbowl party at his house

One night, while aligning the telescope (before the

where Dolores had thrown so many colored paper stars

shelter was built), Dolores was so engaged that when

into the air that three years later, “… I was still finding

advised to go down and warm up she protested. “No,

them when we moved out of the house!” Oh, and her

no, I’m okay. I’m fine. I don’t need to go down,” she

star-studded team won.

recalls saying.

“When I finally did get up, my legs

almost gave out. I was completely numb!”

Dolores is also passionate about the humane treatment of animals, loves to snorkel, camp, and has numerous

Today, Dolores spends a lot less time on the mountain,

cats and (paradoxically) feeds wild birds, which she says,

and while her devotion is physically less straining, her

she can watch forever (as can her cats). Of course, all

commitment to the successful operation of Gemini

her loves—from the stars to her husband (an astronomer

is as strong as ever. “Comparing then to now is like

whom she has been with for more than 20 years), her

comparing a covered wagon to a Mercedes. I never

daughter and two grandchildren, as well as the plants

thought a telescope with so many peripherals could

and animals she protects—have converged on the island

end up working so smoothly,” she added. Of course,

of Hawai‘i, making her life quite complete.

as Gemini astronomer Tom Geballe points out, Dolores is a key reason that this is true. “I’ve known Dolores

As for the future, Dolores confesses that she doesn’t

since our days together at the United Kingdom Infrared

need to work anymore, but continues, “The only reason

Telescope,” he said. “She was the first telescope operator

I’m working is because I enjoy it,” she said. “I’m glad I

hired there and had the same deep involvement as

am where I am.”

she has at Gemini. Her curiosity about all aspects of telescope operations, not to mention astronomy,

Peter Michaud is the Public Information and Outreach Manager

naturally led her to advocate broader responsibilities for

at Gemini Observatory. He can be reached at:

telescope operators than was the norm. This was key to developing the SSA model at Gemini.”

When Dolores looks back at her past 26 years on the mountain, she has myriad memories and unique experiences—from the nighttime visitor who started a fire under the United Kingdom Infrared Telescope slit (which corrupted her data), to natural phenomena like the plasma ball from a lightning storm that seemed to follow her on the Saddle Road; Mauna Kea is a part of Dolores and she, likewise, is a part of the mountain. Her words capture the deep relationship she has established with the mountain. “I feel privileged to be there. I don’t think everyone has a chance to experience that,” she said. “I’ll walk that mountain in pitch darkness. I feel safe on it. It is very unusual. I don’t normally walk in the dark, but I feel that way about the mountain.” So how does a person with Dolores’ overflowing mental and physical energy ever relax? First, she has to be caught up on her e-mail. “I log on… I have to do that before I can actually go and relax!” she admits. “From that point on I’m happy. I love gardening. I love pulling weeds; it’s very relaxing.”



Shadows stretching over Mauna Kea Pu‘u. Mauna Kea’s cinder cones, caught at sunset, with clouds riding the inversion layer beyond. Joy Pollard obtained this image on July 16, 2009, at about 6:30 pm, on the slopes of Mauna Kea outside of the Gemini North telescope. The image was taken during a tour with University of Hawai‘i REU (Research Experiences for Undergraduates) students. Joy used a Nikon D90 with an 18.0 - 105.0 millimeter f/3.5-5.6 lens.



Sunrise at Ahu Tonga Riki, on the west side of Rapa Nui. Manuel Paredes took this photo to create a view of the Moai highlighting the colors of the sunrise as a backdrop. His idea was to show them as part of nature by revealing only their silhouettes so they would not dominate the natural scene. Manuel used a Nikon D1X digital camera with a Nikkor 17 - 35 millimeter f/2.8 lens.



Gemini Observatory Northern Operations Center 670 N. A‘ohoku Place, Hilo, Hawai‘i 96720 USA Phone: (808) 974-2500 Fax: (808) 974-2589 Gemini Observatory Southern Operations Center c/o AURA, Casilla 603 La Serena, Chile Phone 56-51-205-600 Fax: 56-51-205-650

Gemini Partner Agencies:

United States

United Kingdom



Brazil University of Hawai‘i REU students watch a sunset from the altitude platform of Gemini North. Photo by Joy Pollard



The Gemini Observatory is operated by the Association of Universities for Research in Astronomy, Inc., under a cooperative agreement with the National Science Foundation on behalf of the Gemini Partnership. To receive GeminiFocus, please send a request (with your mailing address) to:

Profile for Gemini Observatory

Issue 39 - December 2009  


Issue 39 - December 2009